Skip to main content

Full text of "Journal of South African Botany"

See other formats




suN - h if v 


w\ary. 


Digitized by the Internet Archive 
in 2016 with funding from 

South African National Biodiversity Institute Libraries 


https://archive.org/details/journalofsouthaf50unse 


JOURNAL 

OF 

SOUTH AFRICAN 
BOTANY 

VOLUME 49 1983 

Published under the authority 
of the 

Trustees 
of the 

National Botanic Gardens 
of 

South Africa, 

Kirstenbosch, 

Claremont. 

Editor-in-Chief: 

Prof. H. B. Rycroft (parts 1 , 2 and 3) 

Prof. J. N. Eloff (part 4) 

(Director, National Botanic Gardens of South Africa 

Harold Pearson Professor of Botany, University of Cape Town). 

Editorial Committee: 

Miss L. L. Bosman 
Mr. P. J. Grobler 
Dr. J. P. Rourke 

Dr. D. J. B. Killick (Botanical Research Institute Representative) 

(Mr. E. G. H. Oliver: Alternate) 

Prof. N. Grobbelaar (South African Association of Botanists Representative) 
(Prof. O. A. M. Lewis: Alternate) 

Publication dates: 

49(1): 1-86 [3 January, 1983] _____ 

49(2): 87-180 [1 April, 1983] 

49(3): 181-322 [1 July, 1983] I NAV/ORSlNGSliUStlTuUT VIR PLAnlTKUNDE 

49(4): 323-460 [12 October, 1983] 


BOTANICAL P c SEA n CH INSTITUTE 

DEPT. BF AGRICULTURE AND FISHLRIeT 


JOURNAL 

OF 

SOUTH AFRICAN BOTANY 


VOLUME 49 

CONTENTS 


1983 


Page 


TROPHIC STRUCTURE OF SOME COMMUNITIES 

OF FYNBOS BIRDS. W. R. Siegfried 1 

A LIST OF THE LACHENALIA SPECIES INCLUD- 
ED IN RUDOLF SCHLECHTER’S COLLEC- 
TIONS MADE IN 1891-1898 ON HIS COLLECT- 
ING TRIPS IN SOUTHERN AFRICA, WITH 


IDENTIFICATIONS ADDED. W. F. Barker 45 

FREY UNI A VISSERI EN FREYLINIA DECURRENS 
(SCROPHULARIACEAE): TWEE NUWE SPECIES 
VAN DIE SUID-WES KAAPLAND. E. J. van 
Jaarsveld 57 

A REVIEW OF DECOMPOSITION AND REDUC- 
TION AND OF SOIL ORGANIC MATTER IN 
TROPICAL AFRICAN BIOMES. J. W. Morris 65 


BOOK REVIEWS: Progress in botany 43: morphol- 
ogy, physiology, genetics, taxonomy, geobotany, 
edited by H. Ellenberg, K. Esser, K. Kubitzki, E. 
Schnepf and H. Ziegler (A. R. A. Noel). Biology of 
plants, by P. H. Raven, R. F. Evert and H. Curtis 
(C. H. Bornman). Plant growth regulators: agricul- 
tural uses, by L. G. Nickell (J. van Staden). Sil- 
icon and siliceous structures in biological sys- 
tems, edited by T. L. Simpson and B. E. Volcani 
(F. R. Schoeman). Physiological plant ecology 1: 
responses to the physical environment, edited by 
0. L. Lange, P. S. Nobel, C. B. Osmond and H. 
Ziegler (B. H. Walker). Hormonal regulation of de- 
velopment 1, edited by J. Macmillan (C. H. Born- 
man). Plant growth substances 1979, edited by F. 
Skoog (C. H. Bornman) 


THE DEFINITION OF LEAF CONSISTENCE CAT- 
EGORIES AND THEIR DISTRIBUTION ALONG 
AN ALTITUDINAL GRADIENT IN THE SOUTH 
EASTERN CAPE. R. M. Cowling and B M 
Campbell 


PROTEOID ROOTS IN THE SOUTH AFRICAN 
PROTEACEAE. B. Lamont 


POLLINATION OF MIMETES HIRTUS (PROTE 
CEAE) BY CAPE SUGARBIRDS AND ORANG 
BREASTED SUNBIRDS. B. G. Collins 


NEW XANTHOPARMELIAE (LICHENES) FROM 
SOUTHERN AND CENTRAL AFRICA. M. D. E. 
Knox and F. A. Brusse 


79 


87 

103 

125 

143 


Page 


AN ANNOTATED SYSTEMATIC CHECKLIST OF 
THE ANGIOSPERMAE OF THE CAPE RECEIFE 
NATURE RESERVE, PORT ELIZABETH. M. C. 

Olivier 161 

ALOE MEYERI VAN JAARSVELD, AND A. RICH- 
TERSVELDENSIS VENTER & BEUKES. P. 

Vorster 175 


BOOK REVIEWS: The biochemistry of alkaloids, by 
Trevor Robinson (W. E. Campbell). Basidium and 
basidiocarp: evolution, cytology, function and de- 
velopment, edited by Kenneth Wells and Ellinor 
K. Wells (A. Eicker). Biochemistry and physiology 
of herbicide action, by C. Fedtke (K. H. Schutte). 
Nucleic acids and proteins in plants 1 : structure, 
biochemistry and physiology of proteins, edited 


by D. Boulter and B. Parthier(J. van Staden) 177 

CHROMOSOME NUMBERS IN SOME SOUTH 
AFRICAN SPECIES OF LINUM L. (LINACEAE). 

C. M. Rogers 181 

GALIUM ROURKEI (RUBIACEAE): A NEW 

SPECIES FROM THE SOUTH WESTERN 
CAPE. Ch. Puff 185 

A NEW EUPHORBIA FROM SOUTH WEST AFRI- 
CA. L. C. Leach 189 

A CHECK-LIST OF THE PTERIDOPHYTES OF 

ZIMBABWE. J. E. and S. M. Burrows 193 

A REMARKABLE NEW LEUCOSPERMUM (PRO- 
TEACEAE) FROM THE SOUTHERN CAPE. J. P. 

Rourke 213 

THE TAXONOMIC VALUE OF TRICHOMES IN 
PELARGONIUM L’HERIT. (GERANIACEAE). L. 
Oosthuizen 221 

A NEW SPECIES OF BOOPHANE HERBERT 
(AMARYLLIDACEAE) FROM THE NORTH 
WESTERN CAPE. D. Snijman 243 

NOTES ON TULBAGHIA: 5. SCANNING ELEC- 
TRON MICROSCOPY OF SEED-COAT PAT- 
TERNS IN NINETEEN SPECIES. C. G. Vosa .... 251 

STUDIES IN CYPERACEAE IN SOUTHERN AFRI- 
CA: 10. THE GENUS KYLLINGA ROTTB. F. M. 

Getliffe 261 

TYLECODON KRITZINGERI: A NEW TYLECODON 
(CRASSULACEAE) FROM THE NORTH WEST- 
ERN CAPE. E. J. van Jaarsveld 305 


BOOK REVIEWS: The Monocotyledons: a compara- 
tive study, by Rolf M. T. Dahlgren and H. Trevor 
Clifford (P. Goldblatt). Plant carbohydrates I: in- 
tracellular carbohydrates, edited by F. A. Loewus 
and W. Tanner (N. Grobbelaar). Plant carbohy- 
drates II: extracellular carbohydrates, edited by 
W. Tanner and F. A. Loewus (N. Grobbelaar). 
Growth and differentiation in plants, by P. F. 
Wareing and I. D, J. Phillips (C. H. Bornman). Ex- 
periments in plant tissue culture, by J. H. Dodds 


Page 


and L. W. Roberts (C. H. Bornman). Plant physi- 
ology by I. P. Ting (C. H. Bornman). Plant cell 
and tissue culture, by J. Reinert and M. M. Yeo- 
man (C. H. Bornman). Saline and sodic soils, by 
E. Bresler, B. L. McNeal and D. L. Carter (B. H. 
Walker). Experimental embryology of vascular 
plants, edited by B. M. Johri (J. van Staden). His- 
torical plant geography, by Philip Stott (N. H. G. 
Jacobsen). New perspectives in wood anatorqy, 
edited by Pieter Baas (A. C. Gibson) 

WILHELM JAN LUTJEHARMS (1907-1983). O. A. M. 
Lewis 


TWO NEW SPECIES OF PSORALEA (FABACEAE) 

IN SOUTH AFRICA. C. Stirton 329 

TWO NEW SPECIES OF OTHOLOBIUM (FABA- 
CEAE). C. Stirton 337 

BOWIEA GARIEPENSIS: A NEW BOWIEA 

SPECIES (LILIACEAE) FROM THE NORTH 
WESTERN CAPE. E.J. van Jaarsveld 343 

THE AFRICAN GENUS TRITONIA KER-GAWLER: 

PART 2. SECTIONS SUBCALLOSAE AND 
MONTBRETIA. M. P. de Vos 347 

SIX MORE NEW SPECIES OF LACHENALIA (LIL- 
IACEAE). W. F. Barker 423 

STUDIES IN ORCHIDACEAE FROM SOUTHERN 

AFRICA. G. Williamson 445 

A NEW SPECIES OF ERIOSEMA (FABACEAE) 

FROM THE EASTERN TRANSVAAL. C. Stirton . . 451 


BOOK REVIEWS: Dictionary of cultivated plants and 
their regions of diverstiy: excluding most orna- 
mental forest trees and lower plants, by A. C. Ze- 
ven and J. M. J. de Wet (O. Kerfoot). The study 
of plant structure: principles and selected meth- 
ods, by T. P. O'Brien and M. E. McCully (C. H. 
Bornman). Molecular biology of plant tumours, 
edited by G. Kahl and J. S. Schell (C. H. Born- 
man). Physiological plant ecology: II. Water re- 
lations and carbon assimilation, edited by O. L. 
Lange, P. S. Nobel, C. B. Osmond and H. Ziegler 
(B. H. Walker). Advances in research and tech- 
nology of seeds: part 6, edited by J. R. Thomson 
(J. G. C. Small). Advances in research and tech- 
nology of seeds: part 7, edited by J. R. Thomson 


(J. G. C. Small) 455 

ANNOUNCEMENT 460 

INDEX TO PLANT NAMES: VOL. 49 (1 -4) 461 


Indexed and/or Abstracted in Biological Abstracts, 
Cambridge Scientific Life Sciences Collection, 
Current Advances in Plant Science, Index to 
South African Periodicals, International Bibli- 
ography of Book Reviews and International Bibli- 
ography of Periodical Literature. 



THIS VOLUME IS DEDICATED TO 

Bernard De Winter (1924- ) 

M.Sc., D.Sc. (Pret.) 

(Director of the Botanical Research Institute, Pretoria) 

He joined the Division of Botany and Plant Pathology in 1947 as a research assistant in charge 
of the grass collections of the National Herbarium, Pretoria. In 1950 he was seconded to the 
Royal Botanic Gardens, Kew, as South African liaison officer. In 1959 he was appointed Offic- 
er-in-Charge of Botanical Survey in Pretoria and so began a long and distinguished career in 
botanical administration culminating in his appointment as Director of the Botanical Research 
Institute in 1973. His main contributions to South African botanical literature have been re- 
visions of the South African Ebenaceae, Stipeae, Aristideae and Eragrostis. In all he has col- 
lected nearly 10 000 specimens. He is commemorated in the plant names Kirkia dewinteri, Aloe 
dewinteri , Aristida dewinteri, Silene dewinteri and others. 

He has also been active in the affairs of several scientific societies, namely the South African 
Association for the Advancement of Science, the South African Biological Society and the 
South African Association of Botanists. He has served on the Council of these societies for 
many years and in each case also served a term as President. He was Secretary-General of 
AETFAT from 1978-1982 and played a major part in the organization and ultimate success of 
the 10th AETFAT Congress held in Pretoria in January 1982. 





J1 S. Afr. Bot. 49 (1): 1-43 (1983) 


TROPHIC STRUCTURE OF SOME COMMUNITIES OF FYNBOS 
BIRDS * 

W. R. Siegfried 

( Percy FitzPatrick Institute, University of Cape Town, Private Bag, Ronde- 
bosch 7700, R.S.A.) 

Abstract 

Comparisons are made between aspects of the trophic structure of avian com- 
munities (at the alpha-diversity level) of the four major vegetation types of the fyn- 
bos biome in the south-western Cape, in order to point up some of the roles that 
birds can potentially play in ecological processes in the fynbos ecosystem. More par- 
ticularly, the report deals with selected features of plants as well as birds, in an at- 
tempt to evaluate the probable importance of birds as agents for either pollination or 
seed dispersal in fynbos vegetation. 

Avian standing crop biomass and energy requirements are about five times high- 
er in Coastal Renosterveld and Strandveld than in either Coastal or Mountain Fyn- 
bos, reflecting a major difference in the density of birds. Bird density is not unusually 
low in fynbos generally, in relation to some other terrestrial biome types in southern 
Africa, and the avifauna can be characterized as one of relatively small-bodied birds. 
Primary consumers account for some 50-70 % of the avian-based energy transfer in 
fynbos generally, and secondary and tertiary consumers are almost exclusively insec- 
tivorous. The primary consumer component is greatest in Strandveld and Coastal 
Renosterveld. In Mountain Fynbos the avian primary consumer component ap- 
proaches or slightly exeeeds 70 % apparently only in old vegetation; insectivores be- 
ing relatively more prominent in younger, regenerating stands of vegetation. The 
production of avian food resources is allocated differently in the four main fynbos 
vegetation types and, concomitantly, there are trends for disproportionate increases 
in the relative biomass of frugivores, granivores and, to a lesser extent, folivores with 
an increase in total avian biomass. The nectar-eating class, however, decreases and 
the proportion of the insect-eating class is similar across such an arrangement. More 
particularly, seeds and foliage apparently are more abundantly available to birds in 
Strandveld and Coastal Renosterveld than in Coastal and Mountain Fynbos, and 
fleshy fruits are relatively abundant only in Strandveld. 

Broadly speaking, seasonal cycles of vegetative growth and flowering are oppo- 
site in Strandveld and Coastal Renosterveld on the one hand, and Coastal Fynbos 
and low-altitude Mountain Fynbos on the other. These differences are matched very 
roughly by changes in the seasonal composition of the avifaunas, particularly the nec- 
tarivorous species. However, the changes in avian species composition and abun- 
dance appear to be relatively small in Strandveld and Coastal Renosterveld com- 
munities. In essence, this means that Coastal and Mountain Fynbos avian 
communities are more dynamic, reflecting the birds’ movements in tracking spatially 
and temporarily fluctuating resources. The actual movements made by the birds are 
not well understood, and their study is needed urgently to provide the knowledge es- 
sential for proper conservation of fynbos in the south-western Cape. This applies 
particularly to the nectarivorous birds which apparently are important pollinators of 
many fynbos plants. 


* Accepted for publication 10th March, 1982. 


1 


2 


Journal of South African Botany 


I_J ITT RE ICS EL 

DIE VOEDINGSAMESTELLING VAN SOMMIGE GEMEENSKAPPE FYN- 
BOS VOELS 

Vergelykings word getref tussen aspekte van voedingsamestelling van voelge- 
meenskappe (by die alpha-verskeidenheid vlak) van die vier hoof-plantegroei tipes 
van die fynbos bioom in die suid-westelike Kaap, sodat sommige van die rolle wat 
voels moontlik kan speel in die ekologiese prosesse in die fynbos ekosisteem bepaal 
kan word. Daar word veral gelet op geselekteerde eienskappe van plante en voels, in 
'n poging om bestuiwings- of saadverspreidingsagente in fynbos te bepaal. 

Die biomassa van die voels en hulle energiebehoeftes is sowat vyf keer hoer in 
Kusrenosterveld en Strandveld as in beide Kus- en Bergfynbos wat op ’n groot ver- 
skil in voelbevolking dui. Die digtheid van voels is nie buitengewoon laag in fynbos 
in vergelyking met ander bioom tipes in suidelike Afrika nie en die voelbevolking 
word gekenmerk deur relatief klein voeltjies. Primere verbruikers is sowat 50 tot 
70 % van die voel-energieoordrag in fynbos en sekondere en tersiere verbruikers is 
byna uitsluitlik insekvreters. 

Die primere verbruiker-komponent is die grootste in Strandveld en Kusrenoster- 
veld. In Bergfynbos kom die primere voel verbruiker alleen in ou fynbos naby 70 %. 
Insekvreters is relatief meer prominent in jonger, herstellende stande plantegroei. 
Die produksie van voel voedselbronne is verskillend in die vier hoof fynbos plante- 
groei-tipes en terselfdertyd is daar ’n neiging vir ’n oneweredige vermeerdering in die 
relatiewe biomassa van vrugte-, graan- en, tot ’n mindere mate, blaarvreters met ’n 
toename in die totale voelbiomassa. Die nektaretende klas verminder en die verhou- 
ding van inseketers is insgelyks dwars met so ’n reeling. Dit is veral sade en blare wat 
meer geredelik beskikbaar is vir voels in die Strandveld en Kusrenosterveld as in die 
Kus- en Bergfynbos en vlesige vrugte is slegs in die Strandveld relatief volop. 

In die bree gesien, is die seisoens-groeisiklusse en blomtye teenoorgesteld in die 
Strandveld en Kusrenosterveld aan die een kant en die Kusfynbos en lae hoogte 
Bergfynbos aan die anderkant. Hierdie verskil word rofweg gevolg deur die verande- 
rings in seisoens-samestelling van die voelbevolking, veral die nektareters. Tog blyk 
veranderings in die samestelling van die voelbevolking en voorkoms relatief klein in 
die Strandveld- en Kusrenosterveld-gemeenskappe. Dit beteken in werklikheid dat 
Kus- en Bergfynbos-voelgemeenskappe meer dinamies is omdat hulle bewegings 
tydelik fluktuerende bronne volg. Die werklike beweging van voels word nog nie 
goed verstaan nie en verdere studie is noodsaaklik terwille van die behoorlike bewa- 
ring van die fynbos van die suidwes-Kaap. Dit is veral van toepassing op die nektar- 
eters wat blykbaar belangrike bestuiwers van baie fynbos plante is. 

Introduction 

Each of the four major vegetation types of the fynbos biome in the 
south-western Cape, South Africa, supports a distinct avifauna, with avian 
species richness in communities of Coastal Fynbos, Coastal Renosterveld 
and Strandveld being about double that in Mountain Fynbos (Siegfried and 
C i owe, 1982). Here I report on a preliminary and provisional comparison of 
aspects of the trophic organizations of these avifaunas, in order to point up 
somc the r ^* es ^at the birds potentially can play in ecological processes 
m the fynbos ecosystem. More particularly, I report on selected features of 
plants as well as birds, in an attempt to evaluate the probable importance of 
mds us agents tot either pollination or seed dispersal in fynbos vegetation. 


Trophic structure of some communities of Fynbos birds 


3 


Study Area and Methods 

Detailed descriptions of 11 study sites (Appendix 1) representative of 
major vegetation types in the fynbos biome, and the methods used in sur- 
veying birds and plants in these vegetation types, are given in Siegfried and 
Crowe (1982). Phenological data on flowering, fruiting and vegetative 
growth of 10 individually-marked plants of each of the most prominent (con- 
spicuous and/or abundant) species, in four segments along two transects, 
were recorded each month during at least 12 months in each of the vegeta- 
tion types. The size, form and colour of the species’ seeds, fruits and flowers 
were recorded at each of the 11 study sites. An additional study site com- 
prising 50 ha of mature Coastal Fynbos near Hopefield (S33° 3' E 18° 22') 
was censused regularly for birds, but the vegetation was not studied. 

Results and Discussion 

Avian biomass and consumers 

Avian standing crop biomass in Coastal Renosterveld and Strandveld is 
about five times higher than that in either Coastal or Mountain Fynbos 
(Table 1). The relatively high biomass in Coastal Renosterveld and Strand- 
veld is not due to a disproportionate number of heavier species, but it is 
rather a reflection of denser populations of birds (Fig. 1, Table 2). Avian 
standing crop biomass in Coastal and Mountain Fynbos is considerably low- 
er than that in savanna vegetation (Table 1). However, bird density is not 
unusually low in fynbos in relation to some other terrestrial biome types in 
southern Africa (Table 2). Hence, it is likely that the fynbos avifauna can be 
characterized as one of relatively small-bodied birds. 

Table 1. 

Bird biomass (kg/100 ha) and energy requirement (kcal/m-Vy - ') in savanna and fyn- 
bos vegetation in southern Africa. First approximation annual energy requirement 
was calculated following Tarboton (1980) and using the Lasiewski-Dawson (1967) 
equation. (Comparable figures for other terrestrial biome types in southern Africa 

are not yet available.) 


Vegetation 

Biomass 

Energy 

Source 

Acacia Woodland, Transvaal 

Burkea Woodland, Transvaal 

Strandveld 

Coastal Renosterveld 

Coastal Fynbos 

Mountain Fynbos 

84.6 

40.7 
42,6-45,6 

45,9 
7,9-8, 3 
3,1-14,1 

17,5 

7,5 

9,6-11,2 

10,9 

1.9- 2, 3 

1.9- 2, 7 

Tarboton (1980) 
Tarboton (l980) 
This study 
This study 
This study 
This study 


4 


Journal of South African Botany 


The avian communities of Strandveld and Coastal Renosterveld vegeta- 
tion have energy requirements similar to that of a Burkea savanna avifauna 
(Table 1). Primary consumers account for some 50-70% of the avian-based 
energy transfer in fynbos communities, and secondary and tertiary con- 
sumers are almost exclusively insectivores (Table 3). Primary consumers ap- 


. • MOUNTAIN FYNBOS 
o COASTAL FYNBOS 
" □ RENOSTERVELD 
- ■ STRANDVELD 


.c 

o 

o 


if) 

o 

m 


M 

■ 


1000 


§ 500 


S 

A 



r sMi 


- SM2 


10 20 30 40 
BIRD BIOMASS Ckg/l00ha) 



censused during a full year. 


Trophic structure of some communities of Fynbos birds 


5 


parently rank lower in Acacia and Burkea savanna avifaunas; and, in the 
Acacia community (2,3 x the annual energy requirements of the Burkea 
community) they provide slightly more than 50% of the energy transfer, 
whereas in Burkea it is the secondary and tertiary components that do so 
(Tarboton, 1980). Accepting that the data sets are at best first approxima- 
tions, this trend is paralleled in fynbos in that the primary consumer compo- 
nent is greatest in Strandveld and Coastal Renosterveld which, similarly to 
the Acacia community, occur on soils relatively rich in plant nutrients 
(Kruger, 1979) and whose primary production is suspected to be greater 
than in Mountain or Coastal Fynbos. In Mountain Fynbos, at least, the pri- 
mary consumer component approaches or slightly exceeds 70%, in the avi- 
an-based energy transfer, apparently only in old vegetation; insectivores be- 
ing relatively more prominent in younger, regenerating stands of vegetation. 

There is a paucity of information on standing crop biomass of plant com- 
munities in the different fynbos vegetation types, and virtually nothing is 

Table 2. 

Bird density in vegetation types in southern Africa 


Vegetation 

No. birds/ 
100 ha 

Source 

Valiev Bushveld, eastern Cape 

1 235-3 000 

Winterbottom (1972) 

Thorn Bushveld, eastern Cape 

204 

Skead (1946a) 

Karroid Broken Veld, Klaarstroom 

72-235 

Winterbottom (1972) 

Karroid Broken Veld, Robertson 

764 

Winterbottom (1972) 

Central Upper Karoo 

486 

Winterbottom (1972) 

Karoo, South West Africa 

348 

Winterbottom (1972) 

Floodplain Grassland, Zambia 

1 108 

Winterbottom (1972) 

Floodplain Grassland, Botswana 

1 760 


Themeda Grassveld, eastern Cape 

58-272 

Skead (1946b) 

Southern Tall Grassveld, Transkei 

353 

Winterbottom (1947) 

Cympopogon/ Themeda Grassveld, Lesotho 

299 

Winterbottom (1972) 

Mopane Woodland, Zambezi Valley 

1400 

Winterbottom (1956) 

Brachystegia Woodland, Zambia 

620 

Winterbottom (1972) 

Acacia Woodland, Transvaal 

925-1 600 

Tarboton (1980) 

Burkea Woodland, Transvaal 

399-900 

Tarboton (1980) 

Strandveld 

1 034-1 273 

This study 

Coastal Renosterveld 

1 230 

This study 

Coastal Fynbos 

236-291 

This study 

Mountain Fynbos 

200-431 

This study 

Fynbos 

380-700 

Winterbottom (1978) 


6 


Journal of South African Botany 



Trophic structure of some communities of Fynbos birds 


7 


known about their net primary production. Nevertheless, the large differ- 
ence in avian biomass between Strandveld and Coastal Renosterveld on the 
one hand and fynbos sensu stricto on the other, implies that only a relatively 
small fraction of the primary production in Coastal and Mountain Fynbos 
passes to birds and/or that primary production in these vegetation types is 
relatively low (Fig. 2). There are also indications that production of avian. 


I 

>* 


N, 

CM 


E 

N. 


CO 

o 


> 

O 


c c 

LD 


z 


LD 


Q 

DC 

CO 



10 20 30 40 

BIRD BIOMASS (kg/ 100 ha) 


Fig. 2. 

Relationship between annual energy requirements and standing crop biomass o 
birds in Mountain Fynbos (SL Sir Lowry’s Pass, S Swartboschkloof, SM i Si v ermine, 
L Lebanon), Coastal Fynbos (B Barhuis, H Hopefield), Renosterveld ( e ri 
del), Strandveld (G Geelbek, M Melkbosch) and Burkea savanna vegetation types. 
Based on monthly mean numbers of birds censused during a full year an a a in 

Table 1. 


Journal of South African Botany 


food resources of different quality (e.g. fruits, seeds, foliage, nectar, insects, 
etc.) is allocated differently in the different vegetation types. 

There are indications of disproportionate increases in the relative bio- 
mass of frugivores, granivores and to a much lesser extent folivores conco- 
mitant with an increase in total avian biomass, whereas the nectar-eating 
class decreases and the proportion of the insect-eating group is similar across 
a particular arrangement of vegetation types (Fig. 3). In this context, it is 



i cc t'i \ or n u ^ hi r r P C n r ° P ^ 1 0 ma s s<) gramvorous, frugivorous, nectarivorous and 
(RV) and Strand vel WwT ? (MF) ’ C ° aStal F >' nbos (CF), Renosterveld 

d ix 1 ( ,J } vegetatlon u ty P es - Bird biomass is based on data in Appen- 
dix _ and monthly mean numbers of birds censused during a full year. 


Trophic structure of some communities of Fynbos birds 


9 


likely that seeds and foliage are more abundantly available to birds in 
Strandveld and Coastal Renosterveld than in Coastal and Mountain Fynbos. 
Some Mountain Fynbos plant communities might produce more nectar for 
birds than the communities of any of the three other vegetation types, and 
Mountain and Coastal Fynbos tend to contain relatively high proportions of 
plant species which produce flowers apparently coloured so as to enhance 
their attractiveness to birds (Fig. 4). 


o 

LU 

Q. 

C/) 


I- 

Z 

< 

Q. 


0 C 
O 

CO 





1 

^ 1 ' 1 1 1 

B S 

"1 



o • 


55 


SL 

— 



• 

- 

45 



— 



• MOUNTAIN FYNBOS 
o COASTAL FYNBOS 

M— 

■ 

35 


□ RENOSTERVELD 



■ STRANDVELD 

GD 




a □ 

25 

1 

i i i i i i 

1 - 


25 

35 45 55 



% BWY PLANT SPECIES 


Fig. 4. 


Relationship between the number of plant species having brown, orange or red 
(BOR) flowers and those with either blue, white or yellow (BWY) flowers ra Mo 
tain Fynbos (S Swartboschkloof, SL Sir Lowry’s Pass), Coastal Fy n , os ’ 

Renosterveld (D De Grendel) and Strandveld (G Geelbek M Melkbosch) vege 
tion types. Based on data in Appendix 3. 


10 


Journal of South African Botany 


There are relatively few plants producing fleshy fruits in Coastal and 
Mountain Fynbos. Fruits of this kind are abundant only in Strandveld vege- 
tation (Fig. 5). Species producing fleshy fruits coloured orange/red or pur- 
ple/black, colours presumably attractive to birds, are poorly represented in 
Mountain Fynbos communities (Table 4). Large (>10mm) fruits are un- 
common in fynbos generally, and approximately 70 % of the plant species 
produce small (< 3 mm) fruits mainly as achenes (Table 4). The achenes are 
eaten by a variety of granivorous birds (Milewski, 1978), and at least 17 es- 




1 

n 1 T I r- 

~T 

M 






■ 



25 








G 


- 

CO 



D " 



LU 



□ 


- 

O 






0- 






CO 






1- 

z 

15 


• MOUNTAIN FYNBOS 


— 

< 

_J 



o COASTAL FYNBOS 



a. 


B 

o 

S 

□ RENOSTERVELD 



o* 


■ STRANDVELD 


- 



• 



- 


5 

— 



— 



SL 



- 



• i 

1 . 1 

1 




5 

15 25 

35 




Oy 
' O 

PLANT ABUNDANCE 




Fig. 5. 


Relationship between the number of plant species producing fleshy fruits and the 

bor^WthnTchJf f ,e !f n c d T nd CrOWe ’ 1982 > °f these species in Mountain Fyn- 
terveld m t r i ^ Sir Lowry s Pass), Coastal Fynbos (B Barhuis), Renos- 

(D De Grendel) and Strandveld (G Geelbek, M Melkbosch) vegetation 
types. Based on data in Appendix 3. 


Table 4. 

Percentage plant species with either small (< 3 mm), medium (3-10 mm) or large (>1() mm) fruits, and percentage species with 
either orange/red or purple/black coloured fleshy fruits in fynhos vegetation in the south-western Cape. Fleshy fruits include 
either single-seeded or multi-seeded types; the remaining non-tleshy fruits include pods, nuts, achenes and capsules. Based on 

data in Appendix 3 


Trophic structure of some communities of Fynbos birds 


Z W a 

O c/3 


o ' 

Z 


Qj <L> 


r) ^ (N M f^i 


I o o o o o o o 


I o o o o o o 


(N Os rf iO ' 


1) 
G 03 


OJj-r-T 

<U c 

> I 


C 1> r; 
- C -o C 

! _ Jr _ 

[ ^ "u Kt O « ■ 
; v> u 


» 5 S u * 

O O c3 vh o 
aa 

>,-F Z <u if — i fs co xt- 
S?.S .S C C C C 

ex%CCOOOO 

■E£5E = cccc 

(— c3 ^ ■> > X X -G 
^ ^ ^ rG <u ^ ^ 


So |0 g CQ 
U U S 


12 


Journal of South African Botany 



Aground-foraging insectivores 


T rophic structure of some communities of Fynbos birds 13 

sentially granivorous species occur regularly in fynbos. Acacia and Burkea 
savanna woodland in the Transvaal support similar numbers of granivorous 
species (Tarboton, 1980). Thus, Winterbottom’s (1968a) suggestion that avi- 
an species richness is low in fynbos because the flora is poor in seeds suitable 
for, and available to, birds seems unlikely. Indeed, a relatively fine degree 
of resource partitioning is shown by differences in diet between six sympa- 
tric or partly sympatric Serinus (Fringillidae) species in Mountain Fynbos 
alone (Milewski, 1978). 



Aground-foraging granivores 


Fig. 6. 

Relationship between numbers of insectivorous and granivorous bird species foraging 
either on ground or in vegetation 

O = Mountain Fynbos, 0 = Coastal Fynbos, EH = Renosterveld, 

| = Strandveld 


14 


Journal of South African Botany 


The insectivores tend to contribute the highest number of species to the 
fynbos avifauna, closely followed by granivores (Table 5). Together, insecti- 
vorous and granivorous birds make up more than 75 % of the total species 
richness of the fynbos avifauna. A division of insectivorous and granivorous 
species into ground and above ground (but excluding aerial) foragers shows 
that the relative incidence of ground foragers in both classes is lowest in 
Mountain Fynbos and highest in Coastal Fynbos (Fig. 6). The reason for this 
difference is not clear, but it could be a function of a relatively high seasonal 
incidence of annual herbaceous plants in the lowland vegetation types. 

Seasonality 

Perennial plants apparently grow actively during summer in Mountain 
and Coastal Fynbos, whereas vegetative growth tends to peak during au- 
tumn-winter, following onset of the rainy season, in Strandveld and Coastal 
Renosterveld (Figs 7, 8, 9, 10). Specht (1957) considered a summer peak in 
growth activity to be an outstanding feature of South Australian heaths, as 
compared with Mediterranean garrigue and maquis, and Californian chapar- 
ral which grow in spring when ambient temperatures are less extreme. 

In the relatively fruit-rich Strandveld vegetation both fruiting and flower- 
ing activities tend to peak in the austral spring (Fig. 10). A spring flowering 
peak occurs in comparable vegetation in West and South Australia, Califor- 
nia, Chile and Israel (Mooney et al., 1974). Flowering activity in Coastal 
Fynbos occurs mainly in autumn and winter (Fig. 8). Thus, very broadly 
speaking, seasonal cycles of vegetative growth and flowering are apparently 
opposite in Strandveld and Coastal Renosterveld on the one hand, and 
Coastal Fynbos and also low-altitude Mountain Fynbos on the other hand. 
The plant phenology of Mountain Fynbos is, however, much more heteroge- 
neous than the relatively uniform patterns applicable to the other vegetation 
types. In Mountain Fynbos flowering activity apparently varies in relation to 
both altitude and age of vegetation. Generally in protea-dominated, mature 
vegetation at low altitudes flowering occurs mainly in winter and spring (e.g. 
Swartboschkloof, Fig. 7), whereas in erica-dominated, young vegetation at 
high altitudes flowering tends to be concentrated in summer and autumn 
(e.g. Sir Lowry’s Pass, Fig. 7). 

Species specific, age and altitudinal differences in the flowering phenolo- 
gy of Mountain and Coastal Fynbos communities are reflected by marked 
differences in the seasonal composition of their avifaunas, particularly the 
nectarivorous class. For instance. Cape sugarbirds Promerops cafer are 
abundant only in winter and spring at Swartboschkloof and in Coastal Fyn- 
bos when proteas tend to flower. Orangebreasted sunbirds Nectarinia viola- 
cea are most abundant in autumn at Sir Lowry’s Pass (Figs 11-12) when eri- 
cas tend to flower. 


Trophic structure of some communities of Fynbos birds 


15 


25 


15 


(f) _ 

I- 5 

Z 

< 


< 25 - 

D 
9 
> 

Q 15 

Z 

LU 

O 5 

< 


Z 

LU 

o 

£ 25 

0. 


15 


STEM 


□ 35 


J L 


,1 


I 



lfljL 


FRUIT 


1 



FLOWER 


: 


1 


I 


I 



mjjasondjfmamj 


Fig. 7. 


Vegetative growth, fruiting and flowering phenology in Mountain Fynbos. Shaded 
columns are representative of the Swartboschkloof site (404 plants of - specie 
pled August 1978-June 1979) and the unshaded columns represent Sir Lowry ras 
(543 plants of 61 species sampled June 1978-May 1979). 


16 


Journal of South African Botany 



MJ JASONDJ F M A M J 


.. Fig. 8. 

Vegetative growth, fruiting and flowering in Coastal Fynbos (Barhuis site; 235 plants 
ot 1/ species sampled August 1978-May 1979. 


PERCENTAGE INDIVIDUAL PLANTS 


Trophic structure of some communities of Fynbos birds 


17 


25 

15 

5 


25 

15 

5 


25 

15 

5 


STEM 



MJJ ASONDJ FMAMJ 


Fig. 9. 

Vegetative growth, fruiting and flowering in Renosterveld (De 
plants of 32 species sampled July 1978-May 1979i 


Grendel site; 275 


PERCENTAGE INDIVIDUAL PLANTS 


18 


Journal of South African Botany 



Iriiln.lriiM .■ .1 


S O N D J 


M J 


Fig. 10. 


Vegetative growth fruiting and flowering in Strandveld. Shaded columns are rep- 

— e th° the . G ^ e ' bek . site (41 ° plants of 44 s P ecies sam pled August 1978-June 
dnd the unshaded columns represent Melkbosch (733 plants of 52 species sam- 
pled May 1978-May 1979). 


Trophic structure of some communities of Fynbos birds 


19 


( f) 

□ 

K 

CO 

6 

z 


1 — T 



SUMMER AUTUMN WINTER SPRING 


FlG. 1L , , • . 

Seasonal abundance (mean number of birds in 3 ha) of orangebreaste s un ^ ir s 
(shaded columns) and all other birds (unshaded columns) in a Moun am Y 
community (Sir Lowry’s Pass). Seasons are: summer (December-Fe ruaiy), 

(March-May), winter (June- August) and spring (September- ovem e . 


20 


Journal of South African Botany 


CO 

CO 

< 

5 

O 

CO 


12 - 


CO 

UJ 

o 

UJ 

Q. 

CO 

o 

z 


90 

CO 

Q 

5E 60 

CD 

o 

z 


30 - 


i? 0,5 h 


0,3 - 


0,1 - 


I 


6 



111- 111 n, n. ni n. 


IIiTIiTIi 


I 


0,88 


1 


El 


OS 


J F M A M J J 
MONTHS 


A S 0 N D 


Fig. 12. 

vian species richness (total no. species), mean abundance (no. individuals) and 
mean biomass (kg live weight) according to months in 3-ha areas in Mountain Fyn- 
bos. Shaded columns are representative of the Swartboschkloof site (censuses were 
not made in March, April, May, June and October) and the unshaded columns rep- 
resent Sir Lowry’s Pass. 


Trophic structure of some communities of Fynbos birds 


21 


Seasonal changes in the species composition and abundance of the bird 
communities of the other vegetation types apparently"are less marked (Figs 

13- 15). However, in both Coastal Renosterveld and Strandveld small in- 
creases in the number of species and larger increases in the abundance of 
birds tend to occur during the rainy season between April and October (Figs 

14- 15). These increases coincide with the main growth of annual forbs and 
grasses as well as the perennial plants. While part of the increase in overall 
avian abundance is due to recruitment of young birds following breeding 
which occurs mainly in late winter and early spring (Winterbottom, 1968b), 
it is primarily granivorous, and to a lesser extent insectivorous, birds which 
are involved in the population increase preceding the breeding period. 
Prominent among the granivorous class are the canaries Serinus canicollis , 
S. alario and S. fiaviventris which form wandering, and sometimes large, 
flocks and commonly include soft green seeds of grasses and composites in 
their diets (Winterbottom, 1973; Milewski, 1978). 

A winter maximum in species richness and abundance of granivorous 
and insectivorous birds in lowland fynbos formations has been reported also 
by Winterbottom (1971) who was unable to explain the phenomenon, but 
suggested that food was unimportant as a controlling factor. His opinion ap- 
parently was prompted by a similar suggestion made by Blondel (1969) in 
explaining why the winter avifauna of the Mediterranean garrigue vegeta- 
tion is richer than that of summer in both species diversity and biomass. Part 
of the problem, however, is an artefact caused by a too rigid classification of 
certain avian populations as belonging to resident or sedentary species. In 
the fynbos, at least, it seems likely that a number of so-called residents, or 
sedentary birds, in fact move about extensively, regularly or irregularly, 
within and between major vegetation formations, in response to fluctuating 
food resources. It also seems likely that there are regular movements of 
birds from the Karoo biome into lowland fynbos whose “carrying capacity 
is lowest during the dry, hot summer when many plants are either dormant 
or semi-dormant in Strandveld and Coastal Renosterveld and most annuals 
are dead in Coastal Fynbos as well. 

Leaving aside the swifts (Apodidae) and swallows (Hirundinidae) which 
seldom are encountered foraging low over fynbos, a striking feature of all 
fynbos avifaunas is the virtual absence of seasonal visitors in the form of 
long-distance migrants (Winterbottom, 1971). 


Nectarivores and frugivores 

Four specialized nectarivorous bird species occur in fynbos, althougl 
each species feeds on arthropods as well. In Mountain Fynbos, the ape 
sugarbird feeds mainly on nectar and arthropods associated with proteas 


BIOMASS (kg) NO. BIRDS NO. SPECIES 


22 


Journal of South African Botany 


Avian 
mean I 



J FMAMJ J A S O N D 
MONTHS 


riG. 13. 

immastr^llv? sp , ecies )’ mean abundance (no. individuals) and 

nomass (kg live weight) according to months in a 3-ha area in Coastal Fynbos 
(Barhuis site). 


BIOMASS (kg) NO. BIRDS NO. SPECIES 


Trophic structure of some communities of Fynbos birds 


23 



MONTHS 


Fig. 14. 


Avian species richness (total no. species), mean abundance ( ■ R Pnns t e rveld 

mean biomass (kg live weight) according to months in a 3-ha area 

(De Grendel site). Censuses were not made in Apnl and August. 


BIOMASS (kg) NO. BIRDS NO. SPECIES 


24 


Journal of South African Botany 



J FMAMJJA30ND 
MONTHS 


Fig. 15. 

Avian species richness (total no. species), mean abundance (no. individuals) and 
mean biomass (kg live weight) according to months in 3-ha areas in Strandveld. 
Shaded columns represent the Geelbek site and the unshaded columns represent 

Melkbosch. 


Trophic structure of some communities of Fynbos birds 25 

(Mostert et al., 1980), whereas the orangebreasted sunbird tends to take 
nectar and arthropods associated with ericas (Skead, 1967). Populations of 
these two species are highly mobile, and they apparently move considerable 
distances in tracking spatially and temporarily fluctuating food resources 
(pers. obs.). However, whereas Cape sugarbirds occur regularly in Coastal 
Fynbos, and occasionally in Strandveld, as well as in Mountain Fynbos, 
orangebreasted sunbirds seldom are found outside the latter vegetation type 
(Winterbottom, 1968b; pers. obs.). 

The nectar-arthropod resource base provides for an intermediately-sized, 
third species in the even more itinerant malachite sunbird Nectarinia famosa 
whose populations apparently move widely within and between all four ve- 
getation types of the fynbos. This species may rely to a relatively greater ex- 
tent on nectar supplies associated with ephemeral sources, such as geophytes 
(pers. obs.). The fourth specialist nectarivore in the fynbos is the lesser- 
doublecollared sunbird Nectarinia chalybea which approximates the orange- 
breasted sunbird in size. It is common in the lowland vegetation formation 
where it, too, has highly mobile populations (Schmidt, 1964). It occurs oc- 
casionally in Mountain Fynbos. Thus, Mountain and Coastal Fynbos each 
essentially has three species, of which two (P. cafer and N. famosa ) are 
shared, whereas Strandveld and Coastal Renosterveld each essentially has 
only the same two species (N. chalybea and N. famosa). This difference 
probably mirrors the relative abundance of proteas in Coastal and Mountain 
Fynbos, and the relative abundance of ericas in Mountain Fynbos. 

There are no fruit-eating birds permanently resident in Mountain Fyn- 
bos. The redwinged starling Onychognathus morio occurs itinerantly and 
sporadically, and it eats fruit but probably takes as much food in the form of 
insects and other items. Moreover, it is more properly a bird of mountain- 
ous areas, rather than Mountain Fynbos, and it undertakes long foraging 
journeys to forests and other sources of food. The absence of fruit-eating 
birds in Mountain Fynbos is a consequence of the lack of plants producing 
fleshy fruits; the possibility that non-avian animals are precluding fruit-eat- 
ing birds can be discarded. Similarly, the low incidence of fruit-eating birds 
in Coastal Fynbos reflects a relative paucity of plants with fleshy fruits. This 
probably means that birds are unimportant as dispersers of plants, and that 
long-distance dispersal of plants by birds is not favoured in these two vegeta- 
tion types generally. Many, if not most, of the exceptions are small trees 
found on the edges of streams or forest and/or are forest pioneers. How- 
ever, even they produce either small, leathery fruits (e.g. Heeria argentea, 
Rhus angustifolia, R. tomentosa) or dry capsules (e.g. Kiggelaria africana, 
Maytenus oleoides) which dehisce to display seeds, with colourful arils, to 
birds. A further exception might rest in the fact that Cape sugarbirds trans- 
port and incorporate large numbers of protea seeds in their nests (Table 6, 


26 


Journal of South African Botany 


Burger et al., 1976). However, there is, as yet, no evidence for this being a 
significant form of seed dispersal. 


Table 6. 

Number and mass (g dry wht) of protea seeds (eight spp.) in nests (N = 25) of the 

Cape sugarbird 


No. seeds 

Mass seeds 

Seeds as % of 
total mass of 

x S.D. 

Range 

x S.D. 

Range 

nest 

182 189 

10-785 

2,2 2,5 

0,1-9, 9 

10,6 


It is only in the strand and littoral dune communities of all three vegeta- 
tion types that an enhanced incidence of plants with fleshy fruits is found. 
However, these serai communities are not fynbos proper but exemplify a 
complex interplay of fynbos and Afro-montane forest elements in the suc- 
cession from pioneer littoral to scrub vegetation. Among the dominant 
plants in these communities, Euclea racemosa and Sideroxylon inerme pro- 
duce fleshy fruits and seeds which are dispersed by fruit-eating birds, such as 
the Cape bulbul Pycnonotus capensis and the speckled coly Colius striatus. 
These plants probably have fleshy fruits because directional dispersal of 
their seeds by birds is the most effective way of colonizing areas which satis- 
fy the species’ specific habitat requirements in the form of mesic, sheltered 
sites. 

Broad-leaved fruiting shrubs are common in Strandveld vegetation, nor- 
mally forming relatively tall clumps interspersed with lower, more open, as- 
sociations of restios (Restionaceae), succulents, herbs, dwarf shrubs and 
grasses. Not only are these bush clumps a consequence of localized seed dis- 
persal by birds, but they also help to promote, by providing perches for 
birds, the spread of the introduced Australian Acacia cyclops whose seeds 
also are dispersed by birds (Glyphis et al . , 1981). Prominent among the es- 
sentially frugivorous birds of the Strandveld are the Cape bulbul and the col- 
ies, especially Colius indicus. 

Conclusions 

This report was written in response to a request for information on the 
ecological importance, or otherwise, of birds, which could be useful in de- 
veloping a strategy for more effective conservation of coastal lowland in- 
digenous vegetation in the south-western Cape. I chose to base the report 
on a preliminary and provisional trophic classification of the avifaunas and 
their energy requirements in fynbos communities (at the alpha-diversity 
level), because, in the absence of detailed information on individual species, 
this should provide at leat some initial insight into the ecological roles that 


Trophic structure of some communities of Fynbos birds 


27 


the birds play in these ecosystems. Allied to this, I have included botanical 
information where necessary to reinforce assumptions I make about the im- 
portance of nectarivorous and frugivorous birds in ecological processes af- 
fecting the functioning of fynbos ecosystems. 

In my deliberately broad approach, involving comparisons between the 
avifaunas of the major fynbos vegetation types, I have forfeited many details 
which probably would have emerged had I analysed data for variations in 
trophic structure within community types, as well as between types. In this 
context, Wiens (1981) has shown that caution and scepticism are in order 
when viewing the results of broad comparative analyses of avifaunal com- 
munities, based on single-year surveys over a variety of sites. According to 
Wiens (op cit.): “Rather consistently, ecological patterns are discerned on 
broad geographic or temporal scales of resolution, but when one looks to 
the dynamics of local populations and species assemblages for mechanisms 
to explain them, the patterns are missing, swamped by local variability.” 
This more likely than not applies especially to bird communities in fynbos 
ecosystems which generally are characterized by high levels of ecological 
heterogeneity, on relatively small spatial, and short temporal, scales. I have 
also not considered a number of potentially important matters, such as the 
role of birds as predators on the seeds of plants and, hence, the way that 
this might affect the composition of serai communities of fynbos vegetation. 
Such apparent neglect, in most cases, simply reflects the absence of any 
worthwhile information, but it also indicates the current lack of a sound con- 
ceptual base necessary for beginning an understanding of the factors respon- 
sible for the high level of heterogeneity in the organization of fynbos 
ecosystems. 

In spite of these caveats, two major conclusions are contained in this 
paper which should improve our current understanding of ecological func- 
tioning of fynbos ecosystems and, hence, influence current developments for 
more effective conservation of these ecosystems. First, it seems likely that 
birds are not major participants in energy or nutrient transfers in fynbos 
and, thus, energy fluxes and nutrient cycles in these ecosystems are largely 
the products of ecological interactions between members of other taxa that 
form part of the structure of the ecosystems. Secondly, birds appear to be 
important pollinators in fynbos ecosystems. These two conclusions are am- 
plified below. 

Because information on primary productivity, and the diets of birds, and 
on the abundance of their food resources, is either imperfect or lacking all 
together, it is not yet possible to assess properly the roles of birds in trophic 
energy fluxes and mineral element cycles in fynbos ecosystems. However, 
the estimated annual energy requirements of these bird populations are rela 
tively low. Although this could be taken to mean that much greater amounts 


28 


Journal of South African Botany 


of energy and materials pass through other groups of consumers in fynbos 
communities, it does not mean that birds exert few, if any, significant effects 
on the functioning of fynbos ecosystems. Indeed, the observations now 
available indicate the potential importance of birds, through their patterns 
of food consumption, either as pollinators, especially in Coastal and Moun- 
tain Fynbos, or as seed-dispersers, especially in Strandveld vegetation. 

The information available is too meagre to permit a definitive, quantita- 
tive assessment of the incidence of bird-pollinated plants in Coastal Fynbos 
vegetation. However, the data reported in this paper, and casual obser- 
vations in the field, indicate that at least some Coastal Fynbos communities 
are rich in plant species which are characterized by the “syndrome of orni- 
thophily”. These kinds of plants produce relatively low concentrations of 
nectar and also have visual or morphological floral features typically facili- 
tating avian pollination (Faegri and Pijl, 1979). Although virtually nothing is 
known about the interactions between ornithophilous plants and their polli- 
nators in fynbos, it is probably reasonable to assume that a dynamic com- 
plex of mutual relationships between birds and plants has evolved in re- 
sponse to co-evolutionary selection. Thus, pollination by birds might affect 
inter-specific differences in phenology, dispersion and perhaps speciation in 
Coastal Fynbos plants. This is a subject which merits research. 

The integral functioning of untransformed ecosystems encompassing 
remnant communties of Coastal Fynbos is threatened increasingly by frag- 
mentation and insularization. This already has resulted in the endangerment 
or extinction of numerous plant species (Hall et al, 1980). As immigration 
and visitation rates of avian pollinators to these islands of vegetation de- 
crease, as they probably will under conditions of shrinking resources for 
birds, extinction of plant species must increase and could result in a drastic 
decline in the Coastal Fynbos flora. 

Insularization of Strandveld vegetation is not nearly as far advanced as it 
is in the case of Coastal Fynbos, and especially Coastal Renosterveld which 
has been fragmented and reduced to less than 10 % of its former extent by 
the development of wheatlands (Bigalke, 1979). Moreover, Strandveld gen- 
erally embraces less dynamic heterogeneity, in the form of successional or 
other patchiness of avian habitats, than Coastal Fynbos. Thus, while there is 
a need for a better understanding of the ecological roles of birds, especially 
nectarivorous and frugivorous species, in maintaining Strandveld communi- 
ties, the knowledge is required seemingly less urgently than for Coastal Fyn- 
bos and Renosterveld. 

A final caveat, however, is based on a current paucity of information 
concerning a suspected dependence of populations of nectarivorous and, to 
a lesser extent, frugivorous birds on seasonal occupancy of different vegeta- 
tion types and, hence, the interdependence of plant communities in these 


Trophic structure of some communities of Fynbos birds 29 

vegetation types. In short, while it is known that, for instance, populations 
of the lesser doublecollared sunhird move from one plant community to an- 
other within one vegetation type, it is suspected that they might also move 
between two or more vegetation types. If this is indeed the case, then, a 
priority prerequisite for an effective conservation strategy for any of the 
three major types of coastal lowland, indigenous vegetation must be an hol- 
istic view of the interdependent functioning of the vegetation types in the 
south-western Cape. 

Clearly, there is a case for urgent research aimed at yielding the know- 
ledge and precepts required for optimization of the dispersion, design and 
management of nature reserves intended to conserve fynbos communities. 
Studies of the interrelations of ornithophilous plants and their pollinators 
should constitute a prominent part of this research for it is crucial to know 
the spatial and temporal scales of fynbos habitats, including various succes- 
sional seres and their arrangement, which will ensure the continuing exist- 
ence of populations of avian pollinators and thereby the populations of 
plants dependent on these pollinators. 

Acknowledgements 

Thanks are due to R. K. Brooke, R. A. C. Jensen, S. J. Milton, J. C. 
Sinclair, J. Sommerville and G. D. Underhill for assistance in the field. T. 
M. and A. Crowe and S. R. Williamson helped with the processing of data. 

F. J. Kruger and C. Boucher advised in the initial selection of study sites. P. 

G. H. Frost and A. V. Hall gave helpful advice. The Department of Water 
Affairs, Forestry and Environmental Conservation, and the Divisional 
Council of the Cape, allowed access to study sites in areas under their juris- 
diction. The South African Council for Scientific and Industrial Research, 
through its Fynbos Biome Project, and the University of Cape Town pro- 
vided financial support. 

References 

Bigalke, R. C., 1979. Conservation. In: Day, J., Siegfried, W. R., Louw, G. N. and 
Jarman, J. L. (eds.), Fynbos ecology: a preliminary synthesis: pp. 148-155. 
Blondel, J., 1969. Sedentaire et Migration des Oiseaux dans une Garrigue Mediter- 
raneene. Terre & Vie 3: 269-314. 

Burger, A. E., Siegfried, W. R. and Frost, P. G. H., 1976. Nest-site selection in 
the Cape Sugarbird. Zool. Africana 11: 127-158. 

Faegri, K. and Van der Pijl, L., 1979. The Principles of Pollination Ecology. 3rd 
revised edition. Oxford: Pergamon. 244 p. 

Glyphis, J. P., Milton, S. J. and Siegfried, W. R., 1981. Dispersal of Acacia cyc- 
lops by birds. Oecologia 48: 138-141. 

Hall, A. V., De Winter, M., De Winter. B. and Van Oosterhout, S. A. M., 
1980. Threatened and critically rare plants of southern Africa. Sth. Afr. 
Nat. Sci. Progr. Rep. No. 45. Pretoria: C.S.I.R. 


30 


Journal of South African Botany 


Kruger, F. J., 1979. Plant ecology. In: Day, J., Siegfried, W. R., Louw, G. N. and 
Jarman, M. L. (eds.), Fynbos ecology: a preliminary synthesis: pp. 88-126. 
Sth. Afr. Nat. Sci. Progr. Rep. No. 40. Pretoria: C.S.I.R. 

Milewski, A. V., 1978. Diet of Serinus species in the south-western Cape, with 
special reference to the Protea Seedeater. Ostrich 49 : 174-184. 

Milewski, A. V. and Campbell, B., 1976. Bird diversity in relation to vegetation 
types in the Moremi Wildlife Reserve. Trans. R. Soc. S. Afr. 42 : 173-184. 

Mooney, H. A., Parsons, D. J. and Kummerow, J., 1974. Plant development in 
Mediterranean climates. In: Leith, H. (ed.), Phenology and Seasonality 
Modelling. Vol. 8: 255-267. New York: Springer-Verlag. 

Mostert, D. P., Siegfried, W. R. and Louw, G. N., 1980. Protea nectar and satel- 
lite fauna in relation to the food requirements and pollinating role of the 
Cape sugarbird. S. Afr. J. Sci. 76: 409-412. 

Schmidt, R. K., 1964. The Lesser Double-Collared Sunbird (Cinnyris chalybeus 
(L.)) in the south-western Cape. Ostrich 35: 86-94. 

Siegfried, W. R., and Crowe, T. M., 1982. Distribution and species diversity of 
birds and plants in fynbos vegetation of mediterranean climate-zone South 
Africa. In: Proceedings of the 3rd International Conference on Mediter- 
ranean-type Ecosystems, Stellenbosch 1980. Berlin: Springer-Verlag. 

Skead, C. J., 1946a. A census and study of birds in 2\ acres of Albany open thorn 
bushveld. Ostrich 17 : 213-221. 

Skead. C. J., 1946b. A census and study of birds in Albany open sour grassveld. Os- 
trich 17 : 222-231. 

Skead, C. J., 1967. The Sunbirds of Southern Africa. Cape Town: Balkema. 

Specht, R. L., 1957. Dark Island Heath (Ninety Mile Plain, South Australia) V. The 
water relationships in heath vegetation and pastures on the Makin sand. 
Aust. J. Bot. 5: 151-172. 

Tarboton, W. R., 1980. Avian populations in Transvaal savanna. Proc. IV Pan-Afr. 
Orn. Congr.: pp. 113-124. 

Wiens, J. A., 1981. Single-sample surveys of communities: are the revealed patterns 
real? Am. Nat. 117 : 90-98. 

Winterbottom, J. M., 1947. The bird population of 110 acres in the Transkei. Os- 
trich 18 : 175-178. 

Winterbottom, J. M., 1956. The bird population of mopane woodland. Nth. Rhod. 
J. 3(2): 124-126. 

Winterbottom, J. M., 1968a. Remarks on the possible origin of the avifauna of the 
south west Cape. Ibis 110 : 91-93. 

Winterbottom, J. M., 1968b. A check list of the land and fresh water birds of the 
western Cape Province. Ann. S. Afr. Mus. 53: 1-276. 

Winterbottom, J. M., 1971. Seasonal changes in the passerine avifauna of the Cape 
macchia. Ostrich 42 : 291-293. 

Winterbottom, J. M., 1972. The ecological distribution of birds in southern Africa. 
Monogr. Percy FitzPatrick Inst. Afr. Ornithol. 1 : 1-82. 

Winterbottom, J. M., 1973. Note on the ecology of Serinus species in the Western 
Cape. Ostrich 44 : 31-34. 

Winterbottom, J. M., 1978. Birds. In: Werger, M. J. A. (ed.), Biogeography and 
Ecology of Southern Africa: pp. 949-979. The Hague: Junk. 


Appendix 1 

Plant and bird community study sites and characteristics in the south-western Cape fynbos biome 


Trophic structure of some communities of Fynbos birds 31 



a All 25 year means, except Lebanon (4 yr) and Silvermine (3 yr). 


32 


Journal of South African Botany 


Appendix 2 

Mass (g live wht) and diet classes for avian species. 

1 = raptor, 2 = granivore. 3 = folivore. 4 = insectivore. 5 = scavenger. 
6 = frugivore, 7 = nectarivore. 


Species 

Mass 

Diet class 

Falco biarmicus 

587 

1 

Falco tinnunculus 

130 

1 

Elanus caeruleus 

192 

1 

Circus maurus 

658 

1 

Francolinus africanus 

405 

2 

Francolinus capensis 

750 

2 

Coturnix coturnix 

78 

2 

Turnix hottentotta 

52 

2 

Columba guinea 

350 

2 

Streptopelia capicola 

150 

2 

Stigmatopelia senegalensis 

85 

2 

Oena capensis 

42 

2 

Colius striatus 

52 

3 

Colius colius 

42 

3 

Colius indicus 

58 

6 

Tricholaema leucomelas 

35 

4 

Geocolaptes olivaceus 

170 

4 

Certhilauda albescens 

27 

4 

Mirafra apiata 

29 

4 

Certhilauda curvirostris 

24 

4 

Eremopterix verticalis 

17 

4 

Corvus albus 

700 

5 

Parus afer 

20 

4 

Anthoscopus minutus 

7 

4 

Chaetops frenatus 

41 

4 

Pycnonotus capensis 

40 

6 

Monticola rupestris 

80 

4 

Monticola explorator 

37 

4 

Cercomela familiaris 

21 

4 

Saxicola torquata 

15 

4 

Cossypha caffra 

27 

4 

Erythropygia coryphaeus 

21 

4 

Bradypterus victorini 

10 

4 

Sphenoeacus afer 

31 

4 

Malcorus pectoralis 

11 

4 

Sylvietta rufescens 

11 

4 

Apalis thoracica 

10 

4 

Cisticola fulvicapilla 

9 

4 

Cisticola subruficapilla 

9 

4 

Prinia maculosa 

10 

4 

Parisoma subcaeruleum 

14 

4 

Parisoma layardi 

14 

4 


Trophic structure of some communities of Fynbos birds 


33 


Species 

Mass 

Diet class 

Bradornis infuscatus 

27 

4 

Siegelus silens 

26 

4 

Batis capensis 

12 

4 

Lanius collaris 

41 

4 

Laniarius ferrugineus 

55 

4 

Telophorus zeylonus 

65 

4 

Creatophora cinerea 

70 

4 

Spreo bicolor 

102 

4 

Promerops cafer 

33 

7 

Nectarinia famosa 

17 

7 

Nectarinia violacea 

10 

7 

Nectarinia chalybea 

9 

7 

Zosterops pallidus 

11 

4 

Passer melanurus 

27 

2 

Ploceus capensis 

45 

2 

Euplectes orix 

24 

2 

Euplectes capensis 

20 

2 

Serinus totta 

30 

2 

Serinus canicollis 

15 

2 

Serinus alario 

12 

2 

Serinus sulphurata 

29 

2 

Serinus albogularis 

26 

2 

Serinus flaviventris 

17 

2 

Serinus leucoptera 

24 

2 

Eringillaria capensis 

21 

2 


34 


Journal of South African Botany 


Appendix 3 

List of plant species, fruit and flower types, at study sites in fynbos vegetation in the 
south-western Cape. 

Fruit type: 1 = fleshy with single seed, 2 = fleshy with numerous seeds, 3 = dry, in- 
cluding pods, nuts, achenes and capsules. 

Fruit size: 1 = <3 mm, 2 = 3-10 mm, 3 = >10 mm. 

Fruit colour: 1 = white yellow, 2 = orange red, 3 = purple black, 4 = brown green. 
Flower size: 1 = <3 mm, 2 = 3-10 mm, 3 = >10 mm. 

Flower colour: 1 = brown green, 2 = white yellow, 3 = orange red pink, 4 = blue 
lilac. 

Study sites: 1 = Geelbek, 2 = De Grendel, 3 = Melkbosch, 4 = Swartboschkloof, 5 
= Sir Lowry’s Pass, 6 = Barhuis, 7 = pre-fire Silvermine, 8 = Lebanon 1,9 = Leba- 
non 2, 10 = Lebanon 3, 11 = Lebanon 4, 12 = post-fire Silvermine. 




Fruit 


Flower 

Study 

sites 

Species 

Type 

Size 

Colour 

Size 

Colour 

Adenandra villosa 

3 

1 

4 

2 

2 

5, 12 

Agapanthus africanus . . . 

3 

2 

3 

3 

4 

4 

Agathelpis dithia 

3 

1 

4 

1 

1 

12 

Agathosma bifida 

3 

1 

4 

1 

2 

9, 10, 11 

A. imbricata 

3 

1 

4 

1 

2 

1 

Agyrolobiom sp 

3 

1 

4 

2 

2 

3 

Albuca sp 

3 

2 

4 

2 

2 

1 

A. cooperi 

3 

1 

4 

2 

2 

12 

Amphibolia indecora . . . 

3 

1 

4 

2 

3 

1,3 

Amphithalea sp 

3 

1 

4 

2 

3 

5 

Anagallis arvensis 

3 

1 

4 

2 

2 

1 

Anaxeton asperum 

3 

1 

4 

1 

2 

5 

Anemone capensis 

Anthospermum 

3 

2 

4 

2 

2 

10 

aethiopicum 

3 

1 

4 

1 

1 

1,2,3 

A. bergianum 

3 

1 

4 

1 

1 

5 

A. ciliare 

3 

1 

4 

1 

1 

2 

A. prostration 

3 

1 

4 

1 

1 

8,9, 10 

Antizoma capensis 

1 

2 

2 

1 

2 

1,2,3 

Arctopus echinatus 

3 

1 

4 

2 

2 

2 

Arctotis acaulis 

3 

1 

4 

2 

3 

2 

A. breviscarpa 

3 

2 

4 

3 

3 

1 

Aristea africana 

3 

2 

4 

2 

4 

12 

A. macrocarpa 

3 

2 

4 

2 

4 

12 

A. racemosa 

3 

2 

4 

2 

4 

9 

A. spiralis 

3 

1 

4 

2 

4 

8, 9, 10, 11 

Aspalathus sp 

3 

2 

4 

2 

2 

7, 12 

A. carnosus 

3 

1 

4 

2 

2 

5 

A. ciliaris 

3 

1 

4 

2 

2 

2,5 

A. ericaef olios 

3 

1 

4 

2 

2 

2 

A. hispida 

3 

1 

4 

1 

2 

3 

A. millefolia 

3 

2 

4 

2 

2 

10 


Trophic structure of some communities of Fynbos birds 


35 




Fruit 


Flower 








Study 







Species 

Type 

Size 

Colour 

Size 

Colour 

sites 

Aspalathus retroflexus . . . 

3 

2 

4 

2 

2 

12 

Asparagus capensis 

2 

2 

2 

1 

2 

1,2, 3, 6 

A. comp actus 

2 

2 

2 

1 

2 

2 

A. crispus 

2 

2 

1 

1 

2 

1,2,3 

A. stipulaceus 

2 

2 

— 

1 

2 

1, 3 

A. thunbergianus 

2 

2 

2 

1 

2 

2,4 

Athrixia crinita 

3 

1 

4 

2 

2 

2 

Babiana ambigua 

3 

1 

4 

3 

4 

1, 3 

B. plicata 

3 

1 

4 

3 

4 

2 

B. tubulosa 

3 

1 

4 

3 

3 

3 

Berkheya armata 

3 

1 

4 

3 

2 

2 

B. herbacea 

3 

1 

4 

2 

2 

4 

B. ilicifolia 

3 

1 

4 

3 

2 

5 

Berzelia abrotanoides . . . 

3 

1 

4 

1 

2 

8,9, 10,11 

Blaeria dumosa 

3 

1 

4 

1 

3 

4,8,9, 







10, 11 

B. ericoides 

3 

1 

4 

1 

3 

5 

Bobartia sp 

3 

2 

4 

3 

2 

3 

B. indica 

3 

2 

4 

3 

2 

4, 5 

Brunia alopecuroides . . . 

3 

1 

4 

1 

2 

10 

B. nodiflora 

3 

2 

4 

1 

2 

4, 9, 10 

Bulbine alooides 

3 

1 

4 

2 

2 

2 

Bulbinella triquetra 

3 

1 

4 

2 

2 

2 

Cannamois acuminata . . 

3 

1 

4 

1 

1 

6 

C. virgata 

3 

2 

4 

1 

1 

4 

Carpacoce spermacocea . 

3 

2 

4 

2 

1 

10 

C. vaginellata 

3 

2 

4 

2 

1 

9, 10 

Carpobrotus sp 

2 

3 

4 

3 

2 

5 

Cassytha ciliolata 

1 

2 

1 

1 

2 

4 

Centella virgata 

3 

1 

4 

1 

1 

5 

Cephalophyllum dubium 

3 

1 

4 

3 

2 

3 

Chironia baccifera 

Chondropetalum 

2 

2 

2 

2 

1 

3 

1 

2 

8,9, 10, 11 

deustrum 

3 

2 

4 

C. hookerianum 

3 

1 

4 

1 

1 

5, 8 

Chrysanthemoides 
incana 

1 

2 

3 

2 

2 

1,3 

C. monilifera 

1 

2 

3 

2 

2 

1 

1 

12 

9, 10, 11 

Chrysithrix capensis .... 

3 

1 

4 

1 

1 

C. junciformis 

3 

1 

4 


Cineraria geifolia 

3 

1 

4 

2 

2 

1 

1 

1 

1 

J 

Cliffortia cuneata 

C. graminea 

3 

3 

1 

1 

4 

4 

1 

1 

10, 11 

C. ruscifolia 

3 

1 

4 

1 


C. subsetacea 

3 

1 

4 

1 

J 

Clutia alaternoides 

3 

1 

4 

1 

2 

4, 5 


36 


Journal of South African Botany 




Fruit 


Flower 

Study 

sites 

Species 

Type 

Size 

Colour 

Size 

Colour 

Clutia daphnoides 

3 

2 

4 

1 

2 

1,3,6 

C. polygonoides 

3 

1 

4 

1 

2 

2, 9, 10 

Coleonema juniperinum . 

3 

1 

4 

1 

2 

4 

Colpoon compressum . . . 

1 

3 

3 

1 

2 

1,3 

Commelina sp 

3 

1 

4 

2 

2 

3 

Compositae sp. a 

3 

1 

4 

— 

— 

5 

Compositae sp. b 

3 

1 

4 

2 

— 

1 

Corymbium africanum . . 

3 

1 

4 

2 

2 

5 

C. congestum 

3 

2 

4 

1 

2 

10, 11 

C. enerve 

3 

1 

4 

2 

2 

5 

C. glabrum 

3 

1 

4 

2 

3 

2, 4,5,8, 
9, 10 

Cotula turbinata 

3 

1 

4 

2 

2 

1 

Cotyledon orbiculare . . . 

3 

1 

4 

3 

3 

1,3 

C. paniculata 

3 

1 

4 

3 

3 

1,3 

Crassula capensis 

3 

1 

4 

2 

2 

2 

C. cyrnosa 

3 

1 

4 

1 

2 

1 

C. fulva 

3 

1 

4 

2 

2 

5 

C. muscosa 

3 

1 

4 

1 

2 

1 

C. zeyheriana 

3 

1 

4 

1 

2 

1 

Cymbopogon marginatus 

3 

1 

4 

1 

1 

4 

Cyphia crenata 

3 

1 

4 

2 

2 

3 

C. phyteuma 

3 

1 

4 

2 

2 

2 

C. volubilis 

3 

1 

4 

2 

2 

1 

Diascia unilabiata 

3 

1 

4 

2 

4 

1 

Dilatris corymbosa 

3 

1 

4 

2 

4 

5,7, 12 

Diosma hirsuta 

3 

1 

4 

1 

2 

4,5 

D. oppositifolia 

3 

1 

4 

1 

2 

5 

Diospyros glabra 

1 

2 

3 

1 

2 

4 

Disperis capensis 

Drosanthemum 

3 

1 

4 

3 

4 

5 

calycinum 

3 

1 

4 

2 

— 

2 

Drosera aliciae 

3 

1 

4 

2 

3 

9, 10, 11 

D. hilaris 

3 

1 

4 

2 

3 

8, 9, 10, 11 

D. trinerva 

3 

1 

4 

2 

2 

5 

Ehrharta calycina 

3 

1 

4 

1 

1 

1, 3, 6 

E. dodii 

3 

1 

4 

1 

1 

9 

E. erecta 

3 

1 

4 

1 

1 

1 

E. ramosa 

3 

1 

4 

1 

1 

1, 9, 10, 11 

E. villosa 

3 

1 

4 

1 

1 

6 

Elegia sp 

3 

1 

4 

1 

1 

7 

E. asperifolia 

3 

1 

4 

1 

1 

10, 11 

E. juncea 

3 

1 

4 

1 

1 

5, 9, 10 

E. parvifiora 

3 

1 

4 

1 

1 

5 

E. racemosa 

3 

1 

4 

1 

1 

4, 5,8,9, 
10, 11 


Trophic structure of some communities of Fynbos birds 


37 




Fruit 


Flower 

Study 

sites 

Species 

Type 

Size 

Colour 

Size 

Colour 

Elegia spathacea 

3 

1 

4 

1 

1 

1 

1 

5 

E. stipularis 

3 

2 

4 

7 

E. vaginulata 

3 

2 

4 

1 

1 

9 

Elytropappus sp 

3 

1 

4 

1 

1 

6 

E. rhinocerotis 

Epischoenus 

3 

1 

4 

1 

1 

2 

quadrangularis 

3 

1 

4 

1 

1 

8, 10, 11 

Eragrostis sp 

3 

1 

4 

1 

1 

2 

Eremia totta 

3 

1 

4 

1 

2 

4 

Erepsia pageae 

3 

1 

4 

2 

3 

2 

Erica articularis 

3 

1 

4 

1 

2 

4, 5 

E. calycina 

3 

1 

4 

1 

2 

11 

E. calystegia 

— 

— 

— 

— 

— 

5 

E. cerinthoides 

3 

1 

4 

3 

3 

5 

E. coccinea 

3 

1 

4 

3 

3 

5, 11 

E. corifolia 

3 

1 

4 

1 

3 

5 

E. hispudula 

3 

1 

4 

1 

3 

4, 8, 9, 10, 

11 

E. imbricata 

3 

1 

4 

1 

2 

4, 5 

E. longifolia 

3 

1 

4 

3 

3 

5, 9, 10 

E. lutea 

3 

1 

4 

1 

2 

5, 11 

E. massonii 

3 

1 

4 

2 

3 

5 

E. plukeneti 

3 

1 

4 

3 

3 

4, 5, 9, 10 
5,7 

E. pulchella 

3 

1 

4 

1 

3 

E. sessiliflora 

3 

1 

4 

2 

1 

8, 9, 10 

E. sphaeroidea 

E. transparens 

3 

3 

1 

1 

4 

4 

1 

3 
1 

4 

4 

11 

Eriocephalus africana . . . 
E. racemosa 

3 

3 

1 

1 

4 

4 

1 

1 

2 

1,3 

Eriospermum cernuum . . 

3 

1 

4 

2 

2 

12 

Eroeda imbricata 

3 

1 

4 

2 

2 

5 

Eroedium sp 

3 

1 

4 

1 

3 

5 

Euclea racemosa 

1 

2 

3 

1 

2 

1, 3 

Euphorbia burmanii .... 

3 

1 

4 

1 

2 

1, 3 
1,3 

E. caput-medusae 

3 

1 

4 

2 

3 

E. mauritanica 

3 

1 

4 

2 

2 

1, 3 

Fagelia bituminosa 

3 

2 

4 

2 

2 

12 

Felicia fruticosa 

F. tenella 

3 

3 

1 

1 

4 

4 

2 

2 

4 

4 

1 

2 

1 

Ficinia bracteata 

3 

1 

4 

1 

4 

F. bulbosa 

3 

1 

4 

1 

1 

12 

F. capillaris 

F. composita 

3 

3 

1 

1 

4 

4 

1 

1 

1 

1 

1 

1 

1 

4 

5 

4,5, 12 
1,3,6 
1 5,7,12 

F. deusta 

3 

1 

4 

F. dunensis 

3 

1 

4 

1 

E. filiformis 

3 

1 

4 

1 

1 


38 


Journal of South African Botany 




Fruit 


Flower 

Study 

sites 

Species 

Type 

Size 

Colour 

Size 

Colour 

Ficinia lithosperma 

3 

1 

4 

1 

1 

12 

F. pinguor 

3 

1 

4 

1 

1 

12 

F. tenuifolia 

3 

1 

4 

1 

1 

7 

F. zeyheri 

3 

1 

4 

1 

1 

9, 10, 11 

Galium tomentosum .... 

3 

1 

4 

1 

1 

1,3 

Geissorhiza imbricata . . . 

3 

1 

4 

2 

2 

1 

Geophyte sp. a 

— 

— 

— 

— 

— 

1 

Geophyte sp. b 

— 

— 

— 

— 

— 

1 

Geophyte sp. c 

— 

— 

— 

— 

— 

2 

Geophyte sp. d 

— 

— 

— 

— 

— 

2 

Gerbera asplenifolia .... 

3 

1 

4 

2 

2 

5 

Gnidia humilis 

3 

1 

4 

1 

2 

2 

G. pinifolia 

3 

1 

4 

2 

2 

7 

Grubbia tomentosa 

2 

2 

3 

1 

1 

8,9, 10 

Haemanthus sp 

1 

3 

2 

3 

3 

2 

Hebenstreitia dentata . . . 

3 

1 

4 

1 

2 

1 

H. repens 

3 

1 

4 

1 

2 

1 

Helichrysum sp 

3 

1 

4 

— 

— 

1 

H. crassifolium 

3 

1 

4 

1 

2 

2 

H. cymosum 

3 

1 

4 

1 

2 

2 

H. felinum 

3 

1 

4 

2 

2 

10 

H. indicum 

3 

1 

4 

— 

— 

4 

H. revolutum 

3 

1 

4 

3 

2 

1,2,3 

H. sesamoides 

3 

1 

4 

2 

2 

5,7 

H. teretifolium 

3 

1 

4 

1 

2 

1, 2 

H. vestitum 

3 

1 

4 

3 

2 

5, 7, 12 

Helictotrichon capense . . 

3 

1 

4 

1 

1 

2 

Fleliphila scoparia 

3 

1 

4 

2 

3 

5 

Helipterum canescens . . . 

3 

1 

4 

2 

3 

5 

Hemimeris sabulosa .... 

3 

1 

4 

2 

2 

1 

Herb sp. a 

— 

— 

• — 

— 

— 

3 

Herb sp. b 

— 

— 

— 

— 

— 

2 

Herb sp. c 

— 

— 

— 

— 

— 

2 

Hermannia trifoliata .... 

3 

1 

4 

2 

3 

1,3 

Hermas quinquedentata . 

3 

2 

4 

1 

2 

5 

H. villosa 

3 

2 

4 

1 

2 

5 

Hippia pilosa 

3 

1 

4 

2 

2 

9, 11 

Hypoduscus 

albo-aristatus 

3 

1 

4 

1 

1 

4, 5, 8, 9, 

FI. aristatus 

3 

1 

4 

1 

1 

10, 11 

4, 5, 7, 8, 12 

H. willdenowia 

3 

2 

4 

1 

1 

6, 12 

Ifloga reflexa 

3 

1 

4 

1 

2 

2 

Indigofera heterophylla . 

3 

1 

4 

1 

3 

1,2,3 

I. procumbens 


1 

4 

2 

3 

1 

Klattia partita 

3 

2 

4 

2 

2 

10, 11 


Trophic structure of some communities of Fynbos birds 


39 




Fruit 


Flower 








Study 







Species 

Type 

Size 

Colour 

Size 

Colour 

sites 

Knowltonia hirsuta 

1 

2 

3 

2 

2 

2 

Lachnaea capitata 

Lachnospermum 

3 

1 

4 

2 

2 

6 

umbellatum 

Lampranthus 

3 

1 

4 

4 

2 

2 

9 

12 

promontorii 

3 

1 

2 

2 

Leptocarpus sp 

3 

2 

4 

1 

1 

11 

L. membranaceus 

3 

1 

4 

1 

2 

5, 8, 9, 10 

Leucadendron laureolum 

3 

2 

4 

2 

2 

7, 12 

L. salignum 

3 

2 

3 

3 

2 

5,6,7 

L. sessile 

3 

2 

3 

3 

2 

5 

L. spissifolium 

3 

2 

3 

3 

2 

4,9, 10, 11 

L. xanthoconus 

Leucospermum 

3 

2 

4 

3 

1 

8, 9, 10 
5, 11 

conocarpodendron . . . 

3 

2 

1 

2 

2 

L. oleaefolium 

3 

2 

1 

3 

2 

5 

L. tomentosum 

3 

2 

4 

3 

2 

6 

Leyssera gnaphaloides . . 

3 

1 

4 

2 

2 

i 

Lichtensteinia lacera .... 

3 

2 

4 

1 

2 

4 

Limonium roseum 

3 

1 

4 

2 

3 

1, 3 

Lobelia sp 

3 

1 

4 

2 

4 

7 

L. comosa 

3 

1 

4 

2 

4 

5 

L. coronopifolia 

3 

1 

4 

2 

4 

4, 5, 12 

L. pinifolia 

3 

1 

4 

2 

4 

5 

Lobostemon hispidus . . . 

3 

1 

4 

2 

4 

2 

Lotononis prostrata 

3 

1 

4 

2 

2 

2 

Lycium afrum 

1 

2 

3 

2 

4 

2 

1 

1 

Maytenus heterophylla . . 

1 

1 

1 

2 

2 

M. maritimus 

1 

2 

1 

1 

2 

M. oleoides 

1 

2 

1 

1 

2 

4 

Melandrium undulatum . 

3 

1 

4 

2 

2 

2 

Merxmeullera lanata .... 

3 

1 

4 

1 

1 

12 

M. rufa 

3 

1 

4 

1 

1 

8, 9, 10 

M. stricta 

3 

1 

4 

1 

1 

9 

Metalasia cephalotes .... 

3 

1 

4 

1 

2 

4, 5, 12 
4,5 

M. muricata 

3 

1 

4 

1 

2 

Metzleria humifusa 

3 

1 

4 

1 

3 

5 

1,3 
1,3 
7, 12 
8,9, 10 

Microloma sp 

M. sagittatum 

3 

3 

1 

1 

4 

4 

1 

1 

1 

3 

3 

3 

Microstylis villosa 

Mimetes cucullatus 

3 

3 

1 

2 

4 

4 

L. 

3 

Mohria caffrorum 

— 

— 

— 

1 

1 

1 

1 

z 

1 

2 

Monochlamys sp 

1 

2 

1 

M. albicans 

1 

2 

1 


Montinia caryophyllacea 

3 

2 

4 

1 

1 

2 

4 

Muraltia alba 

3 

1 

4 




40 


Journal of South African Botany 


Species 

Fruit 

Flower 

Study 

Type 

Size 

Colour 

Size 

Colour 

sites 

Muraltia dumosa 

1 

1 3 

| 1 

1 

4 

1 

1 

4 

1 

Myrica kraussinia 

j 1 

1 2 

3 

1 

1 

9.1 l 

Nebelia fragiraoides .... 

3 

1 

i 4 

1 

2 

8 

N. paleacea 

3 

1 

4 

1 

2 

5,9, 10, 11 

Nemesia parviflora 

3 

1 

1 

2 

4 

1 

Nylandtia spinosa 

1 

2 

2 

1 

4 

3,6 

Olea exasperata 

1 

2 

3 

1 

2 

1, 3 

Osmitopsis afra 

3 

1 

4 

2 

2 

9, 10, 11 

O. parviflora 

3 

1 

4 

2 

2 

5 

Osteospermum 







clandestimim 

3 

1 

4 

2 

2 

2 

O. junceum . : 

3 

2 

4 

2 

2 

9 

O. spinosum 

3 

2 

4 

2 

2 

2 

Othonna coronopifolia . . 

3 

1 

4 

2 

2 

3 

O. cylindrica 

3 

1 

4 

2 

2 

1 

O. filicaulis 

3 

1 

4 

1 

4 

1, 3 

O. quinquedentata 

3 

1 

4 

2 

2 

5, 9 

Oxalis sp. a 

3 

1 

4 

2 

2 

3 

Oxalis sp. b 

3 

1 

4 

2 



1 

Oxalis sp. c 

3 

1 

4 





4 

Oxalis sp. d 

3 

1 

4 

2 

2 

2 

Oxalis sp. e 

3 

1 

4 





6 

O. eckloniana 

3 

1 

4 

2 

3 

2 

O. hirta 

3 

1 

4 

2 

3 

2 

0. pes-caprae 

3 

1 

4 

2 

2 

2 

Passerina paludosa 

3 

1 

4 

1 

1 

12 

P. vulgaris 

3 

1 

4 

1 

1 

3, 6 

Pelargonium 







alchemilloides 

3 

1 

4 

2 

2 

2 

P. angulosum 

3 

1 

4 

2 

3 

12 

P. gibbosum 

3 

1 

4 

2 

2 

1 3 

P. labotum 

3 

1 

4 

2 

1 

2 

P. myrrhifolium 

3 

1 

4 

2 

3 

4 

P. saniculaefolium 

3 

1 

4 

2 

3 

4 

P. triste 

3 

1 

4 

2 

1 

2 

Penaea mucronata 

3 

1 

4 

1 

1 

4, 5,7,9, 







10 11 P 

Pentaschistis sp. a 

3 

1 

4 

1 

1 

s 

Pentaschistis sp. b . . . . 

3 

1 

4 

1 

1 

4 7 

P. aristidioides 

3 

1 

4 

1 

1 

12 

P. colorata 

3 

1 

4 

1 

1 

4, 8,9, 

P. curvifolia 

3 

1 

4 

1 

1 

10, 11 
4 

P. palescens 

3 

1 

4 

1 

1 

4 5 

P. tortuosa 

3 

1 

4 

1 

1 

12 

Petalacte sp 

3 

1 

4 

1 

2 

12 


Trophic structure of some communities of Fynbos birds 


41 




Fruit 


Flower 








Study 







Species 

Type 

Size 

Colour 

Size 

Colour 

sites 

Peucedanum ferulaceum 

3 

2 

4 

1 

1 

9, 10 

P. sieberianum 

3 

2 

4 

1 

2 

5 

Phaenocoma prolifera . . 

3 

1 

4 

2 

3 

5 

Pharnaceum lanatum . . . 

3 

1 

4 

1 

1 

3,6 

Phillipia leeana 

3 

1 

4 

1 

1 

5 

Phylica cephalantha .... 

3 

1 

4 

1 

2 

6 

P. ericoides 

3 

1 

4 

1 

2 

3 

P. gracilis 

3 

1 

4 

1 

2 

4,8,9 

P. parviflora 

3 

1 

4 

1 

2 

5 

P. stipularis 

3 

1 

4 

1 

2 

6 

Pinus pinaster 

3 

2 

4 

1 

1 

4 

Polyarrhena reflexa 

3 

1 

4 

2 

2 

9 

Polycarena cephalophora 

3 

1 

4 

2 

4 

1 

Polygala bracteolata .... 

3 

1 

4 

2 

4 

12 

P. myrtiflora 

3 

1 

4 

2 

4 

1 

Polygonum undulatum . . 

3 

2 

4 

1 

2 

2,5 

Priestleya vestita 

3 

1 

4 

2 

2 

5 

Printzia aromatica 

3 

1 

4 

2 

3 

2 

Prismatocarpus fruticosus 

3 

1 

4 

1 

2 

4 

Protea acaulis 

3 

2 

4 

3 

31 

4 , 5 

P. cynaroides 

3 

2 

4 

3 

3 

5,8,9, 
10, 11 

P. neriifolia 

3 

2 

4 

3 

3 

4,5,9 

P. nitida 

3 

2 

4 

3 

2 

4 

P. repens 

3 

2 

4 

3 

3 

4 

P. scabra 

3 

2 

4 

3 

1 

8, 9 

P. speciosa 

3 

2 

4 

3 

3 

9, 11 

Pseudopentameris 




1 

1 


macrantha 

3 

1 

4 

8, 11 

Psoralea aphylla 

Pterocelastrus 

3 

2 

4 

2 

1 

4 

9, 10 

1.3 

1.2.3 

tricuspidatus 

1 

1 

2 

2 

Putterlickia pyracantha . . 

1 

2 

2 

1 

2 

Rafnia perfoliata 

3 

1 

4 

2 

2 

J 

Restiads 

— 

— 

— 

— 


O 

Restio sp 

3 

2 

4 

1 

1 

11 

R. ambiguus 

3 

1 

4 

1 

1 

1 

1 

1 

1 

8, 10, 11 
4,5, 12 
8,9,10,11 
5,8, 10 
3 

R. cuspidata 

R. dispar 

R. egregius 

3 

3 

3 

1 

1 

1 

4 

4 

4 

1 

1 

1 

1 

R. eleocharis 

3 

1 

4 

R. filiformis 

3 

1 

4 

1 

1 

1 

1 

1 

1 

1 

1 

4 , 5 

R. gaudichaudianus .... 

3 

1 

4 


R. monanthus 

3 

1 

4 

8, 10 
10 

R. perplexus 

R. purpurascens 

3 

3 

1 

1 

4 

4 

1 

1 


42 


Journal of South African Botany 




Fruit 


Flower 

Study 

sites 

Species 

Type 

Size 

Colour 

Size 

Colour 

Restio tenuissimus 

3 

1 

4 

1 

1 

5, 7, 12 

R. triteceus 

3 

1 

4 

1 

1 

4, 5, 7. 8, 
9, 10 

Rhus angustifolia 

1 

2 

1 

1 

2 

2,4 

R. glauca 

1 

2 

2 

1 

2 

1,3 

R. incana 

1 

2 

2 

1 

2 

2,3 

R. mucronata 

1 

2 

2 

1 

2 

1,3 

R. rosmarinifolia 

1 

2 

1 

1 

2 

4 

Roella ciliaris 

3 

1 

4 

2 

4 

5, 12 

Ruschia carolii 

3 

1 

4 

2 

4 

3 

Salsola australis 

3 

1 

4 

1 

1 

1 

Saltera sarcocolla 

3 

1 

4 

2 

3 

5 

Salvia africana 

3 

1 

4 

2 

4 

2 

S. aurea 

3 

1 

4 

3 

3 

1, 2, 3 

Schizaea pectinata 

— 

— 

— 

— 

— 

8, 9, 10, 11, 
12 

Scyphogyne muscosa . . . 

3 

1 

4 

1 

3 

5, 12 

Selago spuria 

3 

1 

4 

1 

4 

4, 12 

Senecio sp 

3 

1 

4 

— 

— 

5 

S. aloides 

3 

1 

4 

— 

— 

1 

S. elegans 

3 

1 

4 

2 

4 

1 

S. grandiflorus 

3 

2 

4 

2 

2 

12 

S. lineatus 

3 

1 

4 

2 

2 

2 

S. pinifolius 

3 

1 

4 

2 

2 

2, 9 

S. purpureas 

3 

1 

4 

2 

4 

5 

S. rigidus 

3 

1 

4 

2 

2 

7, 9, 12 

S. umhellatus 

3 

1 

4 

2 

4 

5,9, 10, 11 

Serruria elongata 

3 

1 

4 

2 

3 

5 

Sickmannia radiata 

3 

2 

4 

2 

2 

12 

Solanum sp 

2 

2 

— 

2 

— 

3 

S. guineense 

2 

2 

3 

2 

4 

1,2,3 

Spiloxene capensis 

3 

1 

4 

2 

2 

2 

Staberoha cernua 

3 

1 

4 

1 

1 

4, 5,7,8, 9 
10, 12 

S. distachya 

3 

1 

4 

1 

1 

6 

Stachys aethiopicum .... 

3 

1 

4 

2 

2 

1,2 

S til be vestitum 

3 

1 

4 

1 

2 

12 

Stoebe sp. 

3 

1 

4 

1 

1 

12 

S. aethiopica 

3 

1 

4 

1 

2 

5 

S. cinerea 

3 

1 

4 

1 

1 

4, 9, 10 

S. incana 

3 

1 

4 

1 

1 

7,9, 10, 11 

S. spiralis 

3 

1 

4 

1 

2 

5 

Struthiola hirsuta 

3 

1 

4 

2 

2 

5 

S. linearis 

3 

1 

4 

2 

2 

5 

S. myrsinites 

3 

1 

4 

2 

2 

8 

Sutera hispida 

3 

1 

4 

2 

2 

2 


Trophic structure of some communities of Fynbos birds 


43 




Fruit 


Flower 

Study 

sites 

Species 

Type 

Size 

Colour 

Size 

Colour 

Sympieza articulata 

3 

1 

4 

1 

2 

8 

Synnotia villosa 

3 

1 

4 

3 

2 

2 

Teloschistes capensis .... 

— 

— 

— 

— 

— 

1 

Tetragonia fruticosa .... 

3 

1 

4 

1 

2 

1,2,3 

T. spicata 

3 

2 

4 

1 

2 

2 

Tetraria autumnalis 

3 

1 

4 

1 

1 

4 

T. brevicaiMs 

3 

1 

4 

1 

1 

5, 8, 10, 11 

T. bromoides 

3 

1 

4 

1 

1 

5,8,9, 10 

T. compacta 

3 

1 

4 

1 

1 

5,7 

T. compar 

3 

1 

4 

1 

1 

12 

T. cuspidata 

3 

1 

4 

1 

1 

4, 5 

T. exilis 

3 

1 

4 

1 

1 

8,9, 10, 11 

T. fasciata 

3 

1 

4 

1 

1 

4, 9, 10 

T. fimbriolata 

3 

1 

4 

1 

1 

8, 9, 10 

T. flexuosa 

3 

1 

4 

1 

1 

5,7, 8,9, 
10, 11, 12 

T. thermalis 

3 

1 

4 

1 

1 

4,5,8,11 

T. triangularis 

3 

2 

4 

1 

1 

11 

T. ustulata 

Thamnochortus 

3 

1 

4 

1 

1 

4, 7 

dichotomus 

3 

1 

4 

1 

1 

4,5 

T. fruticosus 

3 

2 

4 

1 

1 

7, 12 

T. gracilis 

3 

1 

4 

1 

1 

5, 8, 9, 
10, 11 

T. obtusus 

3 

1 

4 

1 

1 

6 

T. similis 

3 

2 

4 

1 

1 

9 

T. spicigerus 

3 

1 

4 

1 

1 

1,3 

Themeda triandra 

3 

1 

4 

1 

1 

4 

Thesium sp 

3 

— 

1 

4 

1 



1 

— 

11 

]_ 








J1 S. Afr. Bot. 49 (1): 45-55 (1983) 


A LIST OF THE LACHENALIA SPECIES INCLUDED IN RUDOLF 
SCHLECHTER’S COLLECTIONS MADE IN 1891-1898 ON HIS 
COLLECTING TRIPS IN SOUTHERN AFRICA, WITH 
IDENTIFICATIONS ADDED 

W. F. Barker 

(Bolus Herbarium, University of Cape Town, Private Bag, Rondebosch 
7700, R.S.A. 

Compton Herbarium, Kirstenbosch, Private Bag X7, Claremont 7735, 
R.S.A.) 

Abstract 

A numerical list of Rudolf Schlechter’s Lachenalia collections made during 
1891-1898 on his collecting trips in Southern Africa, with identifications added, has 
been compiled from his duplicate specimens examined in 15 herbaria, to assist in 
bringing all his numerous sets of duplicate specimens in these and other herbaria up 
to date. 

UlTTREKSEL 

'N LYS VAN DIE LACHENALIA EKSEMPLARE IN RUDOLF SCHLECHTER 
SE VERSAMELINGS WAT IN 1891-1898 IN SUIDER-AFRIKA GEMAAK IS, 
MET IDENTIFIKASIES BYGEVOEG 

Die Lachenalia eksemplare deur Rudolf Schlechter versamel gedurende sy versa- 
meltogte in Suider-Afrika vanaf 1891-1898, is chronologies genommer en identifika- 
sies is bygevoeg. Die lys is saamgestel deur sy duplikaat-eksemplare in 15 herbaria te 
ondersoek, sodat sy groot aantal duplikaat-stelle in hierdie, en ander herbaria op da- 
tum kan word. 

Introduction 

Various sets of Rudolf Schlechter’s specimens obtained on his collecting 
trips in Southern Africa during 1891-1898 were distributed to many herba- 
ria, particularly those in South Africa, Europe and Scandinavia. Among 
these collections were a large number of Lachenalia specimens, about half 
of which were either not identified at specific level, or were given specific 
names, some of which were incorrect. To the rest Schlechter had added his 
own manuscript names. A number of the latter have been described or have 
had their manuscript names validated in recent publications, others have 
been correctly identified, while a few have not yet been described or are dif- 
ficult to identify with certainty at present. 

It has therefore been decided to publish an interim list of these speci- 


Accepted for publication 5th July, 1982. 


45 


46 


Journal of South African Botany 


mens, correctly identified where possible, in order to assist in bringing these 
widely distributed collections up to date. 

In compiling this list the “Itinerary of Rudolf Schlechter’s Collecting 
Trips in Southern Africa” by J. P. Jessop, published in the Journal of South 
African Botany 30(3), July, 1964, was used as a basis, and the relevant 
dates, localities, and register numbers from it have been adopted as head- 
ings to each item. 

In addition, Rudolf Schlechter's specimens from the following 15 herba- 
ria were examined and identified: B, BM, BOL, C, G, GRA, K, L, LD, 
PRE, S, SAM, UPS, WU, Z. 

In the list details from Schlechter’s field tickets have been given opposite 
each of his register numbers, which are arranged in numerical order, in the 
left hand column, and the correct or tentative identifications have been 
added below, followed by the symbols of the herbaria in which the speci- 
mens were studied. 

A number of inconsistencies were observed in the labelling, as at times 
dates vary slightly on labels with the same register numbers, at others the 
figures appear to have been transposed on labels with the same dates and 
other particulars. 

In several instances mixed gatherings were found, when two species were 
represented on the same sheet, or two different species on separate sheets 
had been given the same number. In these cases it is advisable to check the 
specimens with the descriptions of the species. 

I wish to express my thanks and appreciation to all Directors and Cura- 
tors of the herbaria mentioned, who made their collections and facilities 
available during visits to their institutions, or sent specimens out on loan. 
Special thanks are due to Professor E. A. Schelpe of the Bolus Herbarium, 
University of Cape Town and to Dr. J. P. Rourke of the Compton Herba- 
rium, Kirstenbosch, for their assistance and advice during my researches on 
the genus Lachenalia. 

References 

Jessop, J. P., 1964. Itinerary of Rudolf Schlechter’s Collecting Trips in Southern Afri- 
ca. Jl S.Afr.Bot 30(3): 129-146. 


Rudolf Schlechter’s Lachenalia Species 

13/6/1892 ? Rondebosch Common (Camp Ground) 

830 Lachenalia 

In arenosis humidis Camp Ground prope Cape Town 
12. VI. 92 R.Schlechter 830 

Lachenalia juncifolia Bak. (Z). 


List of Lachenalia species in Schlechter’s collections with identifications 47 


830 


15-16/7/92 

1189 


29/7/92 

1238 


29/7/92 

1283 


31/7/92 

1251 


27/8/92 

1343 


4/9/92 

1431 


Lachenalia juncifolia Bak. on same sheet 
Lachenalia unifolia Bak. var. (G, Herb. Bois). 


Sir Lowry’s Pass 
Lachenalia 

In clivis arenosis pr. Sir Lowry’s Pass 
15. VII. 1892 R.Schlechter 1189 

Lachenalia unifolia Jacq. var. wrightii Bak. (Z). 


Claremont 

Lachenalia reflexa Thunb. 

Cape Flats pr. Claremont 

29. VII. 1892 R.Schlechter 1238 

Lachenalia reflexa Thunb. (GRA, Z). 


Claremont 

Lachenalia reflexa Thunb. 

Cape Flats pr. Claremont 

29. VII. 1892 R.Schlechter 1283 

Lachenalia reflexa Thunb. (G, S, WU). 


Retreat and Muizenberg 
Lachenalia rubida Jacq. 

In arenos litoralis Muizenberg 
31. VII. 1892 R.Schlechter 1251 

Lachenalia bulbifera (Cyrillo) Hort. ex Asch. & Graeb. (G, 
GRA, S, WU, Z). 

Lion's Mountain 
Lachenalia orchioides Ait. 

In clivis Leonus Cape Town 

27. VIII. 1892 R.Schlechter 1343 

Lachenalia fistulosa Bak. (GRA, PRE, Z). 

New Kloof, Tulbagh and Tulbagh Waterfall 
Lachenalia 

In planitae pr. Tulbagh 

4. IX. 1892 R.Schlechter 1431 

Lachenalia unifolia Jacq. (Z). 


48 


Journal of South African Botany 


17/8/94 

4849 


17/8/94 

4855 


24-25/8/94 

5018 


24-25/8/94 

5024 


24-25/8/94 

5025 


5-10/9/94 

5239 


5-10/9/94 

5274 


Saron (4847-4849 and 4861-4878) 

Lachenalia orchioides Jacq. 

In arenosis prope Saron 

17. VIII. 1894 R.Schlechter 4849 

Lachenalia pallida Ait. (BM, G, SAM, Z). 

Piquetberg Road. Piquetberg Road is now Gouda (4850-4856) 
Lachenalia schlechteri Bak. 

In planitae pr. Piquetberg Road 600' 

17. VIII. 1894 R.Schlechter 4855 

Lachenalia schlechteri Bak. (in Baker’s hand) 

Holotype (Z). 

Olifants River (4989-5031) 

Lachenalia cf. glaucina Jacq. 

In rip. flum. Olifants River 133 m. 

25. VIII. 1894 R.Schlechter 5018 

Lachenalia trichophylla Bak. (BM, C, G, GRA, K. S, UPS. Z). 

Olifants River (4989-5031) 

Lachenalia trichophylla 
Olifants River 400' 

25. VIII. 1894 R.Schlechter 5024 

Lachenalia trichophylla Bak. (GRA). 

Olifants River (4989-5031) 

Lachenalia hirta Thunb. 

In rip. flum. Olifants River 

25. VIII. 1894 R.Schlechter 5025 

Lachenalia hirta Thunb. (C, G, GRA, PRE, S, UPS). 

Piquetberg (5180-5277) 

Lachenalia tricolor Thunb. 

In planitae summi montis Piquetberg 1 800' 

9. IX. 1894 R.Schlechter 5239 

Lachenalia aloides (L.f.) Hort. ex Asch. & Graeb. var. (Z). 

Piquetberg (5180-5277) 

Lachenalia unicolor Jacq. 

In collib. ad flumen Berg River nr Piquetberg 
10. IX. 1894 R.Schlechter 5274 

Lachenalia unicolor Jacq. var. (GRA, Z). 


List of Lachenalia species in Schlechter’s collections with identifications 49 


11/9/94 

5315 


14-15/10/94 

5507 


28-30/6/96 

7937 


5-6/7/96 

8053 


17-21/7/96 

8243 


21/7/96 

8260 


Braakfontein. Locality has not been identified 
Lachenalia 

In collibus pr. Braakfontein 45 m. 

11. IX. 1894 R.Schlechter 5315 

Lachenalia bachmannii Bak. (PRE, Z). 

Houw Hoek and Bot River (5434-5523) 

Lachenalia 

In arenosis Houw Hoek 
15.X. 1894 R.Schlechter 5507 
Lachenalia rosea Andr. (Z). 

Piqueniers Kloof (7928-7961) 

Lachenalia ophiogiossoides Schltr. n. sp. MS. 

Piqueniers Kloof 

28. VI. 1896 R.Schlechter 7937 

Lachenalia sp. indet. cf. L. mutabilis Sweet and L. variegata 
Barker (BM. GRA, K, Z). 

Langekloof and Bulshoek. The Lang Kloof runs NW of Clan- 
william (8029-8050) 

Lachenalia satyrioides Schltr. n. sp. MS. 

Lange Kloof in aridis 500' 

6. VII. 1896 R.Schlechter 8053 

Lachenalia undulata Masson (BM, G, GRA, K, L, PRE, S, 
Z). 

Kareebergen. The farm Kareebergen is about 13 miles SE of 
Bitterfontein (8167-8258) 

Lachenalia versicolor Jacq. 

Karee Bergen 2 000' 

21. VII. 1896 R.Schlechter 8243 

Lachenalia splendida Diels (BM, BOL, G, GRA, K, L, PRE, 
S, Z). 

Klein Vlei (8259-8261) 

Lachenalia pachycaulos Schltr. n. sp. MS. 

Klyn Fontein in saxosis 1 500' 

21. VII. 1896 R.Schlechter 8260 

Lachenalia mutabilis Sweet var. (BM, G, GRA, K, PRE). 


8260 A mixture of L. mutabilis Sweet var. and L. framesii Barker 
(Z). 


50 


Journal of South A frican Botany 


23/7/96 

8315 

Zout Rivier (8314-8318) 

Lachenalia Loeseneriana Schltr. n. sp. MS. 
Zout Rivier in arenosis 1 000' 

28. VII. 1896 R.Schlechter 8315 

8315 

Lachenalia mutabilis Sweet var. (B, BM, BOL, GRA, K, 
PRE, S). 

Lachenalia framesii Barker (G, L, PRE ,Z). 

3/8/96 

8398 

Langekloof (8385-8401) 

Lachenalia ventricosa Schltr. n. sp. MS. 

Langekloof in sabulosis 

3. VIII. 1896 R.Schlechter 8398 

Lachenalia ventricosa Schltr. ex Barker (B, BM, BOL, G, 
GRA, K, PRE, Z). 

12-13/8/96 

Zeekoe Vlei. Zeekoe Vlei is about 9 miles W. of Clanwilliam 

8490 

(8480-8510) 

Lachenalia neglecta Schltr. n. sp. MS. 

Zeekoe Vley in sabulosis 600' 

12. VIII. 1896 R.Schlechter 8490 

Lachenalia neglecta Schlecht. n. sp. MS. (B, BM, BOL, G, 

GRA, K, L, PRE, S, Z). 

A new species? Not yet published. 

25/8/96 

Boontjies Rivier. The Boontjies River joins the Brandewyn 
Rivier about 4 miles ENE of Pakhuis (8663-8678) 

8670 

Lachenalia unicolor Jacq. 

Boontjies River 

25. VIII. 1896 R.Schlechter 8670 

Lachenalia unicolor Jacq. var. (BM, G, GRA, K, L, PRE, Z). 

25/8/96 

8678 

Boontjies Rivier (8663-8678) 

Lachenalia trichophylla Bak. 

Boontjies Rivier 

25. VIII. 1896 R.Schlechter 8678 

Lachenalia trichophylla Bak. (BM, G, GRA, K, L, PRE, S, 
Z). 

26/8/96 

Bidouwberg. The Bidouwberg is N of the Bidouw River 
(8681—8695) 


List of Lachenalia species in Schlechter’s collections with identifications 51 


8687 


26/8/96 

8689 


28-31/8/96 

8725 


28-31/8/96 

8771 


10-11/9/96 

8968 


24-26/11/96 

9403 


Lachenalia monophylla Jacq. (probably meant for L. unifolia 
Jacq.) 

Bidouw Berg 3 600' 

26. VIII. 1896 R.Schlechter 8687 

Lachenalia bolusii Barker (BM, K). 

Bidouwberg (8681-8695) 

Lachenalia monophylla Jacq. (probably meant for L. unifolia 
Jacq.) 

Bidouw Berg 3 600' 

26. VIII. 1896 R.Schlechter 8689 

Lachenalia bolusii Barker (BOL, G, GRA, K, L, PRE, Z). 

Koudeberg near Wupperthal (8718-8787) 

Lachenalia maximiliani Schltr. n. sp. MS. 

Koude Berg in saxosis 

28. VIII. 1896 R.Schlechter 8725 

Lachenalia maximiliani Schltr. ex Barker (BOL, BM, G, 
GRA, K, L, PRE, S, Z). 

Koudeberg near Wupperthal (8718-8787) 

Lachenalia Schwolkeana Schltr. n. sp. MS. 

Koude Berg 

30. VIII. 1896 R.Schlechter 8771 

Lachenalia elegans Barker var. (B, BM, G, GRA, K, L, PRE, 
S, Z). 

Michells Pass (8931-8970) 

Lachenalia pulchella Schltr. n. sp. MS. 

Michells Pass 

11. IX. 1896 R.Schlechter 8968 

Lachenalia unicolor Jacq. 

var. fragrans (Jacq.) Bak. (B, BM, G, GRA, K, L, PRE, S, 
Z). 

Houw Hoek (9374-9435) 

Lachenalia montana Schltr. n. sp. MS. 

Houw Hoek 

24. XI. 1896 R.Schlechter 9403 

Lachenalia montana Schltr. ex Barker (BM, BOL, G, GRA, 
K, L, PRE, S, Z). 


52 


Journal of South African Botany 


28/11/96 

9473 


30/11/96- 

2/12/96 

9530 


10/4/97 

10395 


21/4/97 

10471 


26/4/97 

10549 


Oct. 1896 
10646 


Dunes near Hawston (9459-9480) 

Lachenalia 

Dunes near Hawston 

28. XI. 1896 R.Schlechter 9473 

Lachenalia rosea Andr. (BM, G, GRA, K, L, PRE, S. Z). 

Vogelgat. Vogelgat is to the NE of the Mouth of the Klein 
River (9507-9579) 

Lachenalia 

1 .XII. 1896 R. Schlechter 9530 

Lachenalia salteri Barker (BM. BOL. G. GRA. K, L, PRE, 
S, Z). 

Onrust River (10395-10399) 

Lachenalia rubida Jacq. 

Onrust River 

10.IV.1897 R.Schlechter 10395 

Lachenalia rubida Jacq. (BM, G, GRA, K, L, PRE, S, Z). 
Elim (10471-10476) 

[Lachenalia Kunickiana Schltr. n. sp. ] MS. 

Elim 

21. IV. 1897 R.Schlechter 10471 

Polyxena sp. (B, BM, G, L, S. Z). 

Zeekoevlei. Zeekoevlei is about 6 miles S of Bredasdorp 
(10549-10551) 

Lachenalia pusilla Jacq. 

Zeekoevley 

26. IV. 1897 R.Schlechter 10549 

Lachenalia pusilla Jacq. (BM, G, K, L, SAM, S, Z). 

Saron 

Lachenalia contaminata Bak. 

In collibus prope Saron 

Oct. 1896 R.Schlechter 10646 

Lachenalia contaminata Ait. (BM, G, GRA, K, L, PRE, Z). 

J his number presents some problems, as it comes much later in the 
sequence than the date would indicate. It may possibly have been 
added to the ticket at this later date as its flowering time is during 
September-October and not in July as its numerical position would 
indicate. 


List of Lachenalia species in Schlechter's collections with identifications 53 


4-5/8/97 

10716 


4-5/8/97 

10740 


10-12/8/97 

10816 


13/8/97 

10827 


14/8/97 


10842 


15/8/97 

10856 


Porterville (10714-10745) 

Lachenalia glaucina Jacq. 

In collibus pone Porterville 

4. VIII. 1897 R.Schlechter 10716 

Lachenalia longibracteata Phillips (BM, G, K, L, PRE, S, Z). 

Porterville (10714-10745) 

Lachenalia sessiliflora Schltr. n. sp. MS. 

In collibus pone Porterville 

5. VIII. 1897 R.Schlechter 10740 

Lachenalia sp. indet. cf. L. mutabilis Sweet & L. variegata 
Barker (BM. G. GRA. K, PRE, S, Z). 

Packhuisberg (10793-10819) 

Lachenalia montigena Schltr. n. sp. MS. 

Pakhuisberg 

12. VIII. 1897 R.Schlechter 10816 

Lachenalia elegans Barker var. (BM, G, GRA, K, L, PRE, S, 
Z). 

Brandewyn Rivier (10820-10830) 

Lachenalia ventricosa Schltr. n. sp. MS. 

Brandewyn Rivier 

13. VIII. 1897 R.Schlechter 10827 

Lachenalia ventricosa Schltr. ex Barker (BM, BOL, G, GRA, 
K, L, LD. PRE. S, Z). 

Lammkraal. No Lammkraal has been found but there is a 
Langkraal about seven miles past the Brandewyn Rivier 
(10832-10850) 

Lachenalia violacea Jacq. 

Lammkraal in collibus 

14. VIII. 1897 R.Schlechter 10842 

Lachenalia violacea Jacq. (PRE). 

Agtertuin. Agtertuin is about 4 miles from the Doom River 
on the Clanwilliam side (10851-10868) 

Lachenalia physopus Schltr. n. sp. MS. 

Agtertuin in collibus 

15. VIII. 1897 R.Schlechter 10856 

Lachenalia violacea Jacq. (B, BM, G, GRA, K, L, LD, S, Z). 


54 


Journal of South African Botany 


16/8/97 

10878 


18-19/8/97 

10907 


7-9/9/97 

11160 

11160 

12-13/9/97 

11203 


12-13/9/97 

11223 

19-20/9/97 

11322 


Doornrivier (10869-10884) 

Lachenalia picta Schltr. n. sp. MS. 

Doornfontein in collibus 

16. VIII. 1897 R.Schlechter 10878 

Lachenalia schlechteri Bak. (B, BM, G, GRA, K, L, PRE, S, 
Z). 

Papelfontein. This is probably Papkuilsfontein, about 26 miles 
E of Van Rhynsdorp (10890-10913) 

Lachenalia obscura Schltr. n. sp. MS. 

Papelfontein 

19. VIII. 1897 R.Schlechter 10907 

Lachenalia obscura Schltr. n. sp. MS. (B, BM, G, GRA, K, 
L, PRE, S, Z). 

A new species? Not yet published. 

Brackdam. Brackdam is about 10 miles N of Garies 
(11160-11161) 

Lachenalia sessiliflora Schltr. n. sp. MS. 

Brackdam in collibus 2 000' 

9. IX. 1897 R.Schlechter 11160 

Lachenalia carnosa Bak. (BM, PRE). 

Lachenalia framesii Barker (G, K, L, LD, S, Z). 

I’Aus. Ouss is an old name for Kamieskroon 
(11203-11234) 

Lachenalia uniflora Jacq. (probably L. unifolia Jacq. is meant) 
I’Aus in collibus 2 300' 

12. IX. 1897 R.Schlechter 11203 

Lachenalia schlechteri Bak. (BM, G, GRA, K, L, PRE, S). 

I’Aus. Ouss is an old name for Kamieskroon 
(11203-11234) 

Lachenalia hirta Thunb. 

I’Aus in collibus 2 600' 

13. IX. 1897 R.Schlechter 11223 

Lachenalia hirta Thunb. (BM, G, GRA, K, PRE, Z). 

Concordia (11313-11344) 

Lachenalia concordiana Schltr. n. sp. MS. 

Concordia in collibus 3 100' 

19. IX. 1897 R.Schlechter 11322 

Lachenalia concordiana Schltr. ex Barker (B, BM, G, GRA, 
K, LD, Z). 


List of Lachenalia species in Schlechter’s collections with identifications 55 


21/9/97 Leeuwpoort. Leeuvvpoort is about nine miles N of Concordia 
(11345-11353) 

11366 Lachenalia xerophila Schltr. n. sp. MS. 

Leos Poort in collibus 3 000' 

21. IX. 1897 R.Schlechter 11366 

Lachenalia xerophila Schltr. n. sp. MS. (BM, BOL, G, K, 
LD. Z). 

A new species? Not yet published. 

21/9/97 Goechas. Goechas is a couple of miles N of Leeuwpoort 
(11365-11379) 

11366 Lachenalia 

Goechas in collibus 3 000' 

21. IX. 1897 R.Schlechter 11366 

Lachenalia xerophila Schltr. n. sp. MS. (GRA) 

This locality fits in better into the numbering in Jessop’s list than 
Leeuwpoort in the former. A new species? Not yet published. 

23/9/97 Steinkopf (11381-11383) 

11383 Lachenalia namaquensis Schltr. n. sp. MS. 

Steinkopf in collibus 2 800' 

23. IX. 1897 R.Schlechter 11383 

Lachenalia namaquensis Schltr. ex Barker (B, BM, BOL, G, 
GRA, K, LD. PRE, Z). 

25/9/97 Karoechas (11384-11394) 

11384 Lachenalia polypodantha Schltr. n. sp. MS. 

Karoechas in collibus 3 000' 

25. IX. 1897 R.Schlechter 11384 

Lachenalia polypodantha Schltr. ex Barker (B, BM, GRA, K, 
LD, PRE, Z). 




J1 S. Afr. Bot. 49 (1): 57-64 (1983) 


FREYLINIA VISSERI EN FREYLINIA DECURRENS (SCROPHULARIA- 
CEAE): TWEE NUWE SPESIES VAN DIE SUID-WES KAAPLAND 


E. J. van Jaarsveld 

( Nasionale Botaniese Tuine van Suid-Afrika, Kirstenbosch, Privaatsak X7, 
Claremont 7735, R.S.A.) 

UlTTREKSEL 

Freylinia visseri E. J. van Jaarsveld en Freylinia decurrens Levyns ex E. J. van 
Jaarsveld word beskryf. F. visseri het reeds uitgesterf in die natuur en is deur mnr. 
Visser van totale uitwissing gered. 

Abstract 

FREYLINIA VISSERI AND FREYLINIA DECURRENS (SCROPHULARIA- 
CEAE): TWO NEW SPECIES FROM THE SOUTH WEST CAPE 

Freylinia visseri E. J. van Jaarsveld and Freylinia decurrens Levyns ex E. J. van 
Jaarsveld are described. F. visseri is already extinct in nature but has been saved by 
Mr. Visser from total extinction. 

Freylinia visseri E. J. van Jaarsveld, sp. nov. 

Differt a Freylinia undulata Benth. quia est multo maior, erectus, stolo- 
niformis, multicaulis, ad 3 m altus frutex; foliis linearibus oblanceolatisque, 
recte expansis, leviter revolutis, 20-30 mm longis; corolla purpurea, infundi- 
bulari, 20-25 mm longa, 3-4 mm lata ad basim; lobis effusis; staminodio 
gracili, piloso, errigente e basi corollae tubi; capsula 10 mm longa et 7 mm 
lata. 

Typus: Hermanuskraal, Velddrift (-CD), 20 November 1947, F. G. H. Visser s.n. 
NBG 46515 (NBG, holotypus). 

'n Regopgroeiende veelstammige haarlose immergroen struik tot 3 m 
hoog. Stingels uitlopervormend, lank en reguit en yl vertak aan basis; ouer 
stingels blaarloos, grysbruin en roedevormig; jonger stingels gerib as gevolg 
van die blaarbasis wat oorlopend op die stingels is. Blare regopgroeiend- 
spreidend, afwisselend of soms in kranse, half-sittend of gewoonlik met n 
kort blaarsteel 1-2 mm lank; lineer omgekeer eiervormig, toegespits en 
gemukroneer aan punt en wigvormig aan basis, 15-30 mm lank en 3-5 mm 


Vir publikasie aanvaar 23 Julie, 1982. 


57 


58 


Journal of South African Botany 


breed, gaafrandig en effens revoluut, liggroen; hoofaar prominent, meer so 
uitstandig aan ondervlak; syare onduidelik. Bloeiwyse eindstandige veel- 
blommende raseem 30-60 mm lank of ’n los pluim tot 190 mm lank. Blom- 
me in 1-3 blommende kortgesteelde byskerms 3-4 mm lank; byskerms af- 
wisselend, teenoorstaande of soms in kranse van 3; skutblare 2-4 mm lank, 
soos die loof blare. Bloeisteel 3-4 mm lank. Kelk 4 mm lank, segmente oor- 
vleuelend, eiervormig en skerppuntig. Blomkroon pers-kleurig 20-25 mm 
lank. Kroonbuis tregtervormig 17 mm lank, 2-4 mm breed aan basis en ver- 
wyd gelydelik na 4-5 mm waar buis 5-lobbig eindig, effens donsharig aan 
die binnevlak; lobbe rond ongeveer 5 mm lank, spreidend. Meeldrade 4 di- 
dinamies, ingebed in boonste helfte van kroonbuis; helmdrade oorlopend op 
kroonbuis, helmknoppe ongeveer 1 mm in deursnee, stuifmeel roomkleurig. 
Staminodium ongeveer 10 mm lank, donsharig en ingebed aan die basis van 
die kroonbuis. Stamper 15 mm lank, speldekopvormig. Vrugbeginsel elip- 
soied 2-sellig, 1,5 m lank. Kapsule ongeveer 10 mm lank en 7 mm in deur- 
snee. 

Blomtyd: September tot November 
Habitat en Verspreiding 

Hierdie plant het uitgesterf in die natuur maar het oorspronklik voorge- 
kom in die Strandveld van die westelike Kusstrook (Acocks Veldtipe Nr. 
34) op die plaas Hermanuskraal (tans tussen Weglopersheuwel en Markus- 
kraal) ongeveer 10 km vanaf die kus. Die plante is vernoem na mnr. Floors 
Visser, wat die plant in November 1947 versamel het. Volgens mnr. Visser, 
het die plant net op daardie plaas voorgekom, op ’n oorblywende “eieland- 
jie” in die Koring-saailand. Die grond was te swak om te verbou as gevolg 
van die onderliggende potklei onder die sand, en daarom dat hierdie lappie 
grond beskermend gebly het. Die lappie grond was ook effens hoer as die 
saailand. Later, in 1954, is hierdie lappie grond ook verbou, waarby hy 4 
lote op Grootklipfontein (Aurora) kom plant het, en wat hierdie spesie van 
uitwissing gered het. 

Volgens mnr. Visser, het die volgende plante ook in assosiasie met F. 
visseri voorgekom: Putterlickia pyracantha, Tetragonia fruticosa, Salvia afri- 
cana, Euphorbia mauritanica en ’n Restionaceae spesie. 

Die skrywer en mnr. Graham Duncan het die gebied by Velddrif baie 
goed gefynkam en daar kon geen spoor van nog populasies gevind word nie. 
Reenval in die gebied is ongeveer 300 mm per jaar. 

Bespreking 

Freylinia visseri is verwant aan Freylinia undulata Benth. maar kan mak- 
lik onderskei word deur die groter en meer regop groeiwyse (3 m). Die 


Freylinia visseri en Freylinia decurrens: twee nuwe spesies 


59 



GPS (L>. 

Fig. 1. 

Map showing distribution of Freylinia species. 
0 F. visseri 
| F. decurrens 


blare is lynvormig omgekeer eiervormig en 20-30 mm lank. Die blomkroon 
is 20-25 mm lank en 3-4 mm breed aan die basis en die blomtyd is van Sep- 
tember tot November. In ’n nota aangeheg aan mnr. Visser se eksemplaar, 
lewer wyle Professor Compton die volgende kommentaar: 


60 


Journal of South African Botany 



Fig. 2. 


Freylinia visseri, sp. nov. NBG 46515 (NBG), 1. bloeiwyse; 2. blomkroon; 3. lengte- 
snee van die blomkroon; 4. stamper; 5. vrugkapsule. 


Freylinia visseri en Freylinia decurrens: twee nuwe spesies 61 

“This seems to be Freylinia related to Freylinia undulata. A striking fea- 
ture is the slender pilose staminode rising from the very base of the co- 
rolla tube; this not being found in the flowers of Freylinia undulata, F. 
lanceolata and F. decurrens" . 

Die plante het plaaslik as Suurlat bekendgestaan. Gelukkig kweek die 
plante maklik van steggies en uitlopers en is reeds by Kirstenbosch Bota- 
niese Tuin gevestig. Hierdie spesie is ’n baie aantreklike tuinstruik en be- 
hoort goed in moeilike sanderige tuine na aan die kus in suid-Kaapland te 
aard. 

Eksemplare Ondersoek 

KAAPLAND — Ex Hort: Kirstenbosch Botaniese Tuin 213/82, E. J. van Jaarsveld 
5677 (Mnr. Visser se tuin by Grootklipfontein Aurora, 19/9/80). (NBG). 


Freylinia decurrens Levyns ex E. J. van Jaarsveld, sp. nov. 

Differt a Freylinia visseri van Jaarsveld foliis distinctis coreaceis canalicu- 
latis et effusis subrecurvatis. Frutex glaber, 1-2 m altus; foliis verticalibus 
quinquefariis, magis aggregatis in extremitatibus ramorum; corolla purpu- 
rea, infundibulari, 16 mm longa; lobis albis revolutis recurvis; capsula 
10-12 mm longa et circa 5 mm lata. 

Typus: Suid van Ladismith (-CB), Julie 1928, M. R. Levyns 2729 (BOL, holotypus). 

’n Regopgroeiend veelstammige haarlose immergroen struik, 1-2 m 
hoog. Stingels uitlopervormend, bruingrys, vertak; jonger stingels ligbruin 
vierkantig, gerib as gevolg van die blaarbasis wat oorlopend op die stingels 
is. Blare liggroen, meer gedronge aan die punte van die takke, in 5 vertikale 
rye, kransgewys of afwisselend, leeragtig, spreidend, lineerlansetvormig, on- 
derskeidend gekanaliseer en teruggekrom, skerppuntig en gemukroneer aan 
die punt, wigvormig aan basis 12-30 mm lank en 2-3(4) mm in deursnee, 
halfsittend of kort gesteeld 1-2 mm, hoofaar prominent, meer so aan onder- 
vlak. Bloeiwyse eindstandig; veelblommende raseem 30-60 mm lank, of ’n 
pluim tot 150 mm lank. Blomme in 1-3 blommende kort-gesteelde (2 mm) 
byskerms. Skutblare geleidelik oorlopend in loofblare, blomsteel 2 mm 
lank. Kelk 5 mm lank, segmente oorvleulend eiervormig, skerppuntig en 5 
mm lank. Blomkroon 16 mm lank, perskleurig, lobbe wit. Kroonbuis treg- 
tervormig, 12 mm lank, silindries, 2,5 mm breed aan basis en verwyd gelei- 
delik na 4 mm, donsharig aan basis van binnevlak, kroonlobbe teruggekrom 
en revoluut, rond, ongeveer 5 mm lank. Meeldrade 4 didinamies, ingebed in 
boonste helfte van kroonbuis, helmdrade oorlopend op kroonbuis; helm- 


62 


Journal of South African Botany 



XO mm 

Fig. 3. 


Freylinia decurrens, sp. nov. E. J. van Jaarsveld 5769 vanaf Pampoenkloof, Montagu 
(NBG). 1. die stingel met die bloeiwyse; 2, die kapsule; 3. die saad. 


Freylinia visseri en Freylinia decurrens: twee nuwe spesies 


63 


knoppe ongeveer 1 mm in deursnee, stuifmeel roomkleurig. Stamper 12 mm 
lank; stigma speldekopvormend. Vrugbeginsel ellipsoied 2-sellig, 1 mm 
lank. Kapsule ongeveer 10-12 mm lank en 5 mm breed. Saad orbikuler met 
membraanagtige vlerk, 2-3 mm in deursnee. 

Blomtyd: Junie tot September. 


Habitat en Verspreiding 

Freylinia decurrens kom voor in die Karoo-agtige gebroke veld (Acocks 
Veldtipe Nr. 26) en die habitat is droog. Reenval kom hoofsaaklik in die 
winter voor en wissel van 200-300 mm p.j. Die plante kom in bergagtige, 
goedgedreineerde gebiede voor, gewoonlik aan suidelike hellings tussen 
rotse en klippe. 

Die skrywer het plante versamel in 'n rotsagtige, skalieryke omgewing. 
’n Enkel plant, in dieselfde omgewing, is ook in 'n stroombedding versamel, 
waar dit saam met F. lanceolata gegroei het (Pampoenkloof, Montagu). 


Bespreking 

Freylinia decurrens is verwant aan F. visseri en F. undulata maar is mak- 
lik onderskeibaar deur die leeragtige, lineerlansentvormige, gekanaliseerde 
teruggekromde spreidende blaartjies en die teruggekromde kroonlobbe. 

Freylinia decurrens is eerste versamel in Julie 1928 deur mev. Levyns wat 
dit as 'n nuwe spesie herken het en dit toe die manuskripnaam F. decurrens 
toegeken het. Dit is ’n aantreklike tuinstruik wat maklik van steggies kweek 
en behoort goed in droe tuine in winterreenvalgebeide aan te pas. 


Eksemplare Ondersoek 

KAAPLAND — 3321 (Ladismith): Kruywagendsdrift, Klein Swartberg (-AC), 1 Julie 
1959, J. M. Wurtz 1670 (BOL). 

—3320 (Montagu): Baden Hillside (-CA), 22nd September 1946, R. H. Compton 
1834 (NBG). 


Bedankings 

Mnr. Winter, Kurator van Kirstenbosch Botaniese Tuin, wat dit vir my 
moontlik gemaak het om die plante na te spoor. 

Dr. Rourke en mevrou Fairall van die Compton Herbarium word vir hul 
hulp bedank. 

Mnr. Norval Geldenhuis word vir die latynse beskrywing bedank en 
mnr. Grobler en mej. Bosman vir nasien van die manuskrip. 

Mev. Ward-Hilhorst word bedank vir die puik illustrasies. 


64 


Journal of South African Botany 


Literatuurverwysings 

Acocks, J. P. H., 1975. Veld types of South Africa. 2nd ed. Mem. bot. Surv. S.Afr. 

40 . 

Dyer, R. A., 1959. Freylinia tropica. In: Flower PI. Afr. Plate 1320. 

, 1975. The Genera of Southern African Flowering Plants. Vol. 1. Pretoria: 

Government Printer. 

Hiern, W. P., 1904. Scrophulariaceae. In: Thiselton-Dyer, Flora Capensis. Vol. 4. 
London: Reeve. 


J1 S. Afr. Bot. 49 (1): 65-78 (1983) 


A REVIEW OF DECOMPOSITION AND REDUCTION AND OF SOIL 
ORGANIC MATTER IN TROPICAL AFRICAN BIOMES 

J. W. Morris 

(Datametrical Services, Department of Agriculture, Private Bag X116, Pre- 
toria 0001, R.S.A.) 

Abstract 

Decompostion and reduction processes are described in relation to the chief 
causative organisms, namely, termites, ants, earthworms, dung beetles and micro-or- 
ganisms while the effects of both inter- and intra-seasonal droughts on these pro- 
cesses and that of fire as a reducer agent are also discussed. Finally, a review of the 
humus status of tropical and subtropical soils, and non-forest ones in particular, is 
presented with particular reference to the possible reasons for the low soil organic 
matter contents generally found in such soils. It is concluded that further research in 
this field is required. 

UlTTREKSEL 

’N OORSIG VAN DIE AFBRAAK EN REDUKSIE EN VAN GROND ORGA- 
NIESE INHOUD VAN TROPIESE BIOME IN AFRIKA 

Afbraak en reduksie prosesse word beskryf in verhouding tot die hoof oorsaak- 
like organismes, naamlik, termiete, miere, erdwurms, miskruiers en mikro-orga- 
nismes terwyl die invloed van beide inter- en intraseisoens-droogtes op hierdie pro- 
sesse sowel as vuur as 'n reduseerder ook beskryf word. Ten slotte, word die 
humusstatus van tropiese en sub-tropiese gronde, en nie-woud gedeeltes veral, aan- 
gebied met spesiale verwysing na die moontlike oorsake vir die lae grond organiese 
materiaalinhoud wat algemeen aangetref word in sulke gronde. Die gevolgtrekking 
word gemaak dat verdere navorsing in hierdie veld nodig is. 

Introduction 

In tropical and subtropical biomes, there are three possible fates for pri- 
mary production that dies before being consumed by an animal. It may be 
consumed when dead (dead matter consumption and reduction), be decayed 
by micro-organisms (decomposition; also referred to as “oxidation ) or be 
burnt. After a brief consideration of litter input, the processes of reduction 
and decomposition are dealt with in this review while herbivory (consump- 
tion) and fire are discussed to a much lesser extent. Dead organic matter 
plays a major role in determining the structure and function of an ecosystem 
by acting as a nutrient reservoir for intrasystem cycling and as an energy 
source for heterotrophic organisms (Singh & Gupta, 1977). Rates of decom- 
position and reduction and consequently the rate of nutrient release, eluci- 


Accepted for publication 16th August, 1982. 

65 


66 


Journal of South African Botany 


dation of which is the usual reason for such studies, depend on the species 
concerned and abiotic factors of the environment as will be pointed out be- 
low. Aspects of the build-up of humus, an end-product of these processes, is 
an enigma in tropical and subtropical biomes and is therefore given promi- 
nence in a section of this review. 

Species of termites, ants, earthworms, dung beetles, Coleopteran larvae, 
millipedes and cockroaches are the macrofaunal components identified as 
making the greatest contribution to dead matter breakdown and consump- 
tion to tropical and subtropical biomes. The relative importance of these 
groups varies from one biome type to another. Microbial activity is at least 
as important as the macrofaunal elements in the decomposition and reduc- 
tion process and under tropical non-forest conditions has major differences 
in mode of functioning from both temperate and tropical forest ecosystems. 
Aspects of these differences are elaborated below. An outline of the pro- 
cesses involved in dead matter consumption and decomposition is also 
given. 

In view of the paucity of data, a very wide definition of tropical and sub- 
tropical biomes is adopted although the geographical coverage of the review 
is heavily biased towards Africa. More emphasis is placed on botanical as- 
pects than zoological although the subject is truly multidisciplinary, a reflec- 
tion on the author’s background. 


Litter Dynamics 

In general modelling terms, inputs to the decomposition and reduction 
system are independent of the standing crop of dead organic material and 
outputs are proportional to, and therefore dependent on, the amount of the 
standing crop. The net rate of change in energy or material stored in an eco- 
logical system, or its parts, equals the rate of input minus the rate of loss 
(Olson, 1963). From one-year litter bag trials at the South African Savanna 
Ecosystem Project study site, Nylsvley, in the Northern Transvaal, Morris et 
al. (1982) reported the following half lives (time for the decomposition of 
half a given amount of litter) for freshly-fallen tree leaves: Ochna pulchra 
5,5 years, Burkea africana 5,1 years and Terminalia sericea 2,9 years. Vege- 
tative material of the two common grasses, Eragrostis pallens and Digitaria 
eriantha, both had half lives of 0,7 years. Values of k, the annual decompo- 
sition constant, for Nylsvley range from 1,20 for grass litter to 0,13 for Och- 
na pulchra leaves in litter bags. The decomposition constant for tree leaf lit- 
ter at Nylsvley is between 0,13 and 0,33. Much faster rates are recorded by 
Malaisse et al. (1975): in miombo woodland, k varies from 1,11 to 1,32 (the 
latter in the presence of fire) and k is 1,40 in dry evergreen forest. Rates 
from 2,0 to 4,85 are quoted for mixed dry lowland forest. In Senegal, Bille 


Decomposition and reduction of soil organic matter in African biomes 67 

(1978) records values for the monthly rate of disappearance of dead matter 
ranging from 0,37 to 0,64 (at the end of the rainy season) to 0,09 to 0,15 
(during the dry season). Where an annual burn is part of the environment, 
such as at Lamto (Cote d'Ivoire) such calculations are not appropriate and 
values are not available. Ashton (1975) records half lives of 0,82 and 1,98 
for leaf and total litter, respectively, from Eucalyptus regnans forest near 
Melbourne, Australia. 

Processes 

Generalisations about dead matter consumption and decomposition pro- 
cesses in tropical and subtropical biomes are difficult to make as the range of 
patterns within such biome types is almost as great as the range between 
them. The best data available on these patterns are from Lamto, a moist, 
rather atypical “savanna” 1 adjacent to tropical rainforest (Lamotte, 1977 
and 1978; Anon, 1979a). At Lamto, dead matter consumption and decom- 
position are largely due to the activities of termites, earthworms and micro- 
arthropods (Lamotte, 1975 and 1979). Ants are also involved, as well as 
myriapods and the underground larvae of Coleoptera. It seems that drier bi- 
omes, as exemplified by Nylsvley, show a markedly different pattern with 
reducers being more important relative to decomposers, termites of great 
importance and earthworms unimportant (Morris et al. , 1982; Huntley & 
Morris, 1982). At Lamto, animals account for nearly 15% of the mineralis- 
ation of organic matter while the mechanical action of these populations 
(soil movement and organic matter transformation) seem even more import- 
ant than mineralisation (Anon, 1979a). The individual roles of important 
agents are discussed below. 

Of the key ecological processes operating in tropical and subtropical bi- 
omes, the inter-seasonal soil moisture deficit and fire have certain, probably 
rather minor, influences on dead matter consumption and decomposition, 
making these processes rather different from those in other biomes. In all 
but the most moist of such biomes, the intra-seasonal intermittent desic- 
cation of litter in the wet season, on the other hand, may have considerable 
ramifications on the processes. 

Termites 

Termites are inhabitants of tropical, subtropical, semi-arid and, to some 
extent, warm temperate regions (Allison, 1973; Wood, 1976). The vast ma- 


1 On the advice of a referee, the term “savanna” has been avoided except whci. 
used in quoted literature as the term in the African context was unanimously recom- 
mended as being unacceptable to botanists and ecologists as long ago as 1966 and the 
resolution has been endorsed by many scientists down ensuing years. 


68 


Journal of South African Botany 


jority of research on decomposition processes has, however, been done in 
cool temperate regions (Anderson & Macfadyen, 1976) and although a few 
species of termites occur in these regions they have little impact on decom- 
position processes. Where termites are abundant, their activities are suf- 
ficient to generate new pathways of decomposition, in comparison with tem- 
perate regions (Wood, 1976). According to Russell (1961), termites in 
tropical soils seem to be the predominant animals affecting the soil and the 
plant. The role of humus-feeding species in pedogenesis is to produce, 
through their faeces, a more stable organic matter substrate for the physio- 
chemical and bacterial agents of degradation (Malaisse et al . , 1975). 

On the basis of diet, there are three kinds of soil-inhabiting termites: 
wood feeders, fungus growers and humus feeders (Harris, 1955). At Lamto, 
foraging, humivore and fungus-growing termites are the three main groups 
(Lamotte, 1975). Harvester termites are an important ecological group 
which feed on living and dead grass while other termites either forage for or 
live in dead wood, or even living trees. Species differ greatly in their diet, 
specificity and habitats and most typical non-forest biome sites have a 
diverse termite fauna. At least 20 species, for example, have been identified 
at Nylsvley (Ferrar, 1982a & 1982b) and 36 species are recorded from Lam- 
to (Wood & Sands, 1978). 

The food of termites consists of plant material of all kinds, including 
wood, grasses, herbs and roots, in the range from living plants to decom- 
posed plant remains mixed in mineral soil. Soil-feeding has been adopted as 
a specialised habit by many species of higher termites. In tropical non-forest 
regions, where levels of organic matter are notoriously low (Wood, 1976), 
termites may be partly responsible for reducing organic matter content. Cel- 
lulose is the main carbohydrate utilised by termites that feed on wood or 
vegetable matter. Digestion in most species is brought about by flagellates 
or bacteria that live in the intestines and secrete cellulase. The five families 
of lower termites (about 25% of known species) have flagellates while the 
higher termites (Termitidae) have no flagellates but do have bacteria 
(P. Ferrar, pers. comm.) The species that do utilise organic residues do so very 
completely and markedly hasten the decomposition process (Allison, 1973). 

There are many common termite species which move large quantities of 
soil, sometimes ingesting much of what they move, bringing grass and other 
plant material into the soil and burrowing and tunnelling through the soil in 
the process (Russell, 1961). Unlike other soil animals which deposit faeces 
within the soil where they are available to micro-organisms and copropha- 
gous invertebrates, termites use their excreta to construct certain regions of 
the nest or to construct fungus combs which are further utilised as food 
(Wood & Sands, 1978). Thus organic material and the nutrients it contains 
are collected from a wide area and concentrated largely in the central region 


Decomposition and reduction of soil organic matter in African biomes 69 

of the nest system. Through the concentration of nutrients in nests, termites 
contribute to the patchiness of nutrient distribution in soils. This process has 
an impact on the vegetation, resulting in greater species diversity and 
stronger patterns (Malaisse, 1978). Trapnell et al. (1976) report the lack of.) 
significant change in nitrogen and carbon in soils under various fire regimes 
in Zambia. After 23 years without burning the organic matter content had 
not changed appreciably. At the same time they report that termite mound 
material had over three times the organic carbon content of adjacent surface 
soil. A proportionally still greater increase in exchangeable bases is report- 
ed. Wood- and litter-consuming species are considered responsible for the 
limitation of humus sources under fire protection. Lee & Wood (1971) and 
Wood & Sands (1978) also report that some termite mounds have concen- 
trations of organic matter higher that the soil from which they are con- 
structed. 

The distinction has, however, to be made here between the humus-feed- 
ing species which build mounds with an excreted paste and the wood- and 
litter-feeding large-mound builders which construct their foraging passages 
and spires with soil particles carried up from subsoil far below the surface 
and bound by salivary matter (Trapnell et al., 1976). Analyses of mound ma- 
terial with organic matter and nutrient contents lower than adjacent top soil 
but closely similar to the subsoil are summarised by Wood & Sands (1978). 
Subsoil is brought to the surface from a depth of from one to more than 
three metres (Allison, 1973). The large mounds that may reach a height of 
10 metres provide microclimates and shelters for a wide range of plants and 
animals within the “savanna” matrix (Malaisse & Anastassiou-Soquet, 1977). 

Ants 

In South America, ants replace termites in some of the scavenging and 
reduction roles that the latter occupy in Africa. Data are available from 
Surinam, Brazil and other tropical South American countries. It is reported 
by Bucher (1982) that ants are of considerable importance in the Chaco as a 
result of their great abundance and degree of ecological diversification. 
Some species of leaf-cutting ants are specialised as detritus feeders, particu- 
larly taking fallen leaves and insect frass. They are extremely abundant in 
the Chaco dry woodlands and as they carry litter underground for fungus- 
growing, they also have an important role in nutrient cycling and redistribu- 
tion. Ants are a relatively minor contributor to soil formation and organic 
matter decomposition according to Allison (1973). He considers that their 
chief role in soil formation lies in the transport of subsoil to the sui tace. 

The role of ants in ecosystems is reviewed by Petal (1978). Roles similar 
to those described above for termites are reported although they appear to 
occupy a wider range of niches than termites. Accumulation of nutrients an 


70 


Journal of South African Botany 


organic matter in nests as well as the opposite effect are found (Petal, 1978). 
Description of the important role of ants as consumers is beyond the scope 
of this review. 

Earthworms 

Earthworms are intolerant of drought and frost and hence dry sandy soils 
and thin soils overlying rock are not usually favourable environments for 
them (Anon, 1979a). They also need reasonably aerated soil and can only 
flourish in soils well provided with organic matter (Russell, 1961). As con- 
ditions in drier tropical and subtropical biomes are not often suitable for 
earthworms, it is not surprising that they are seldom common except in 
moist habitats, such as found at Lamto, where soil organic matter content is 
low (J. C. Menaut, pers. comm.). Some soil-dwelling termites appear, in 
fact, to be the tropical analogue of the earthworm (Russell, 1961). 

Dung beetles 

The main group of dung beetles involved with the reduction of dung is 
the Scarabaeinae (Coleoptera), which are only active in fresh dung. The at- 
tack on dried dung by termites appears to be a characteristic feature of trop- 
ical and subtropical non-forest biomes, judging from evidence from South 
America, Lamto and Nylsvley. Dung beetles attack dung in three main pat- 
terns, of which the best known is perhaps by beetles fashioning lumps of 
fresh dung into spherical balls and rolling them away some distance for buri- 
al (Bornemissza, 1979). More common, but less well-known, are those 
which bury fresh dung on the spot by digging tunnels which may end in 
chambers. The third, and insignificant group, excavate old dung pads. 

The lack of an indigenous Coprophagous fauna in the Australian hinter- 
land for the efficient treatment of cattle dung is worth noting. Although 
there are 250 species of dung beetle in Australia, very few find the soft cow 
pads with their high water content attractive (Bornemissza, 1979). The prob- 
lems arising from dung pollution in the absence of these organisms in Aus- 
tralia is a good example of the essential role decomposers and reducers have 
in a system, whether natural or artificial. 

Micro-organisms 

The complex relationships between groups of micro-organisms and 
specific environmental factors under tropical conditions are reviewed by 
Mohr et al. (1972). Moisture, access of air, temperature, acidity and the sup- 
ply of food materials have different but rather marked influences on the 
number of micro-organisms as well as their activity. Generally, conditions 
appear to favour bacteria in the lower warmer tropics while fungi predomi- 
nate in cooler climates found at higher altitudes. 


Decomposition and reduction of soil organic matter in African biomes 71 


Fire 

In non-forest tropical and subtropical biomes, fire is an important, if not 
the most important, agent of dead matter “respiration”, in comparison with 
its minor influences on consumption and decomposition. The subject of fire 
in these biomes is beyond the scope of this review and only some aspects di- 
rectly related to decomposition and reduction will be mentioned. 

The major effect of fire is the removal of living material (which is poten- 
tial input to the dead organic matter compartment), standing dead plant 
parts and litter and the associated release of organically bound mineral ele- 
ments in the form of ash. Fire will also kill a proportion of the organisms in 
the remaining litter and will make post-fire soil and litter temperatures more 
extreme. Blackened surfaces will absorb more heat while reduced shading of 
the surface by plants and litter will result in raised maximum day-time and 
lowered minimum night-time temperatures. Decomposer organisms can be 
expected to recover from fire fairly quickly while reducers, through a de- 
crease in the quantity and quality of their food source, will recover more 
slowly. 

If fire is excluded from a vegetation type in which fire is normally a ma- 
jor reducer, there will be a build-up of dead organic material until the 
steady state standing crop is reached. The steady state is approached asymp- 
totically and the final amount may be considerable. Such a system has rele- 
vance to the frequency of fire in that a vegetation prone to fire will have a 
high primary production and either a low herbivory rate or a low dead mat- 
ter consumption and decomposition rate, or both. 

Effects of Soil Drought 

Inter-seasonal drying 

Inter-seasonal drying is likely to have effects that are either too broad or 
too subtle to be easily identified. In wetter biomes, organisms such as earth- 
worms are able to survive such drying whereas in drier biomes, dry periods 
within the wet season (see below) will have marked effects confounding the 
effects of the dry season. Micro-organisms are unlikely to be inhibited dur- 
ing the wet season by the necessity to survive the dry season in spore form. 
There may, however, be certain effects on the timing of nutrient release at- 
tributable to drying out of litter, possibly enhancing leaching at the begin- 
ning of the subsequent wet season. 

Intra-seasonal drying 

Considerable effects may result from the frequent dry periods during the 
wet season, during which litter dries out to a great extent. This is a major 


72 


Journal of South African Botany 


contributory factor to the slow decomposition rate recorded for non-forest 
tropical biomes. Micro-organisms are not only inactive in dry litter but have 
to build up in numbers when the litter is re-moistened. This pulse activity 
may itself influence the micro-organism species present as they will usually 
be present in small numbers within the litter and will therefore rarely be 
competing with each other. The wetting-drying cycle within a season may 
have appreciable effects on nutrient release not observed in a continuously- 
moist litter. If nutrients, and especially nitrogen, are more easily washed out 
of dead micro-organisms than from living ones or dead plant matter, then 
there would be a pulse of nitrogen leached at the start of each rainy period 
and rain season. 

The slow rate of decomposition in drier tropical and subtropical biomes 
is likely to have a large number of effects on dead matter consumption and 
decomposers. Just as secondary production in different biomes has been 
characterised as reducer- or consumer-dominated, dead matter consumption 
and decomposition in drier biomes is reducer-dominated with a relatively 
minor role played by decomposers. The opposite holds in moist biomes, re- 
sulting there in low standing crops of litter. In drier biomes, reducer organ- 
isms, with superior water conservation relative to micro-organisms, are able 
to survive dry periods and play an important role in dead matter consump- 
tion and decomposition. Termites have a particularly effective water conser- 
vation system. In moister ecosystems, the termite moisture control system 
(tunnels built for foraging) becomes a handicap as it requires considerable 
energy to maintain. Other, more mobile reducers, take over and decompos- 
ers will be better able to decay material. 

Soil Humus Status 

The relatively slow decomposition rate noted for dry tropical and sub- 
tropical biomes makes it difficult to explain the documented absence of a 
well-defined humus layer in the soils of drier biomes (Anon, 1978; Green- 
land & Nye, 1959; Vine, 1968). In moist biomes including tropical and sub- 
tropical forest, high soil and litter moisture contents and high temperatures 
result in the rapid breakdown of dead organic material and the formation of 
a distinct humus layer. In drier biomes, one or both of these two factors may 
be inoperative while, of course, primary productivity and thus input to the 
litter compartment are also lower. Although the breakup of litter is rather 
slow, once it has broken up and is in small fragments, either on the soil sur- 
face or mixed into the topmost soil layer, the moisture conditions of the ma- 
terial become much more favourable for rapid decomposition. Fresher ma- 
terial on top of small fragments protects the latter from desiccation to a 
certain degree and soil will usually have far better moisture retaining pro- 


Decomposition and reduction of soil organic matter in African biomes 73 

perties than litter so that once the fragments have become incorporated into 
the soil, decomposition is facilitated. The process is further accelerated by 
high temperatures relative to temperate region ecosystems. It is also sug- 
gested that leaching and runoff following tropical downpours could account 
for the absence of a humus layer. Light humus particles on the soil surface 
could also be blown away by wind erosion. Some apparent contradictions 
and gaps in our knowledge with regard to soil humus status became appar- 
ent during the writing of this review, these being presented below with the 
intention that they spur further research. Both Anon (1979a) and Rodin & 
Bazilevich (1967) stress that very little research has been undertaken to date 
in this important field, with particular reference to tropical and subtropical 
areas. 

Soil humus is of great importance in the soil-formation process and in 
soil fertility maintenance (Anon, 1979b). Mohr et al. (1972) consider soil or- 
ganic matter to be the dominant factor in soil formation. As well as being a 
valuable source of nutrients, particularly nitrogen, the organic matter in 
tropical and subtropical biome soils is important in maintaining their struc- 
ture, water-holding capacity and resistance to erosion and in providing much 
of their cation exchange capacity (Jones, 1973). Organic matter influences 
physical and chemical properties of soils far out of proportion to the small 
quantities present, accounting, for example, for at least half their cation ex- 
change capacity (Brady, 1974). Within the tropics, soil organic matter has 
been shown to increase with increasing altitude, increasing rainfall and de- 
creasing temperature (Jones, 1973). Nye & Greenland (1960) note that in 
tropical soils, decreasing rainfall is correlated with lower organic matter con- 
tent. The optimal conditions for mineralisation of plant residues are found 
where temperatures are high and soil moisture satisfactory, such as in the 
low hilly regions of the tropics, and consequently little organic matter will 
remain in its original or humified state (Mohr et al., 1972). On the other 
hand, at higher elevations organic matter will accumulate; the higher the 
elevation the more humus remains (Mohr et al., 1972). 

The formation of humic substances is due to complex transformations of 
the original organic residues and the process and products are described as 
follows by Kononova (1961). In forming humus, the enzymatic activity of 
both micro-organisms and macrofaunal elements are important. A first 
group of substances in soil humus, derived from decomposing plant and ani- 
mal residues and being products of their decomposition and re-synthesis in 
micro-organisms, consists of various nitrogenous and non-nitrogenous oi- 
ganic compounds belonging to well-known groups of organic chemical com- 
pounds such as proteins, carbohydrates, organic acids, fats, waxes and 
resins. A second and larger fraction consists of humic substances. Because 
of the peculiarity of their nature they cannot be related to any existing 


74 


Journal of South African Botany 


groups of organic chemical compounds. Their nature, origin and properties 
are not yet fully understood. Schnitzer & Khan (1972) state that the mode of 
formation of humic substances has been the subject of much speculation and 
list four hypotheses which have been proposed for their synthesis. They con- 
clude that it is difficult to state which of the four hypotheses (plant alter- 
ation, chemical polymerisation, cell autolysis and microbial synthesis) is the 
more valid one. 

Greenland & Nye (1959) have found that decomposition constants for 
lowland (tropical) forest soils are about three times larger than those of non- 
forest tropical biome soils. They consider that the low rate of decomposition 
in the latter soils is probably connected with the repressive effect which 
grasses exert on humus mineralisation, as shown by the very low levels of 
nitrate in such soils. For comparison, constants for temperate-zone oak and 
pine forests have humus-carbon decomposition constants closely similar to 
those for highland tropical forests. The loss of organic matter caused by high 
rates of respiration at high soil temperatures is not always the primary cause 
of low organic matter contents in tropical soils (Greenland & Nye, 1959). 
The effects of low returns of organic matter due to burning and soil erosion 
are also very important in their estimation. Nye & Greenland (1960) note 
that burning in tropical biomes reduces the buildup of humus in the topsoil 
by reducing the amount of litter supplied to the soil organisms. The organic 
matter content of the surface horizons of non-forest soils is considerably 
lower than that of forest soils because of the reduced additions of organic 
matter from the vegetation. The exchange capacity is consequently lower 
(Nye & Greenland, 1960). 

Greenland & Nye (1959) have pointed out that in the tropics the re- 
lationship between soil organic matter and vegetation type is the reverse of 
what is found in the temperate zone (see also Jenny, 1950; Greenland & 
Kowal, 1960; Sanchez, 1976). In the tropics, much the largest addition of or- 
ganic matter to the soil takes place under forest vegetation and the soil or- 
ganic levels tend to be greater than under non-forest vegetation types. In the 
temperate zone, on the other hand, soil organic matter production tends to 
be greater on grassland areas and it is the grassland soils which have the 
highest levels of soil organic matter. A factor seldom considered in such 
general comparisons is that input from primary production should include 
both above- and below-ground plant components, where the latter are very 
difficult to quantify. 

The organic matter content of certain tropical soils, especially the forest- 
ed ones, is low, according to Allison (1973), primarily because of the high 
temperatures and rainfall, abundant plant and animal life in the soil and the 
scarcity of inorganic colloids with which the organic matter might combine. 
Tropical soils can nevertheless support large amounts of vegetation per unit 


Decomposition and reduction of soil organic matter in African biomes 75 

area because moisture and temperature are suitable for growth while roots 
bring up nutrients from the subsoil and also intercept nutrients that are be- 
ing carried down by rainwater. 

Allison (1973) also states that organic matter decomposes at a slower 
rate in the presence of clay than in sand. According to Sanchez (1976), in 
general, the higher the clay contents, the lower will be the annual decompo- 
sition rate of soil organic carbon. This finding is supported by Jones (1973) 
who also notes that, in general, amounts of organic matter and nitrogen in 
the surface soils of the West African “savanna” are small. The mean carbon 
content of 605 well-drained sites was 0,68%. Two important factors govern- 
ing amounts of organic matter in well-drained soils appear to be the clay 
content and a moisture factor related to the length of the wet season (Jones, 
1973). Multiple linear regression on soil clay content and rainfall accounted 
for 46,5% and 57,2%, respectively, of the observed variability of soil carbon 
and nitrogen contents. These findings suggest that the low levels of organic 
matter in “savanna” soils arise from their predominantly sandy nature and 
from the relatively low rainfall. 

Jones (1973) states that soil organic matter status is undoubtedly closely 
related to the rate of return of organic residues to the soil and, hence, the 
type of vegetation. Both vegetative growth and soil microbiological activity 
are closely controlled by soil moisture. He suggests, therefore, that the 
length of the rainy season is more important in determining organic matter 
status than annual rainfall and notes that the essential factor embracing both 
rainfall and drainage factors is the annual period during which the soil is 
moist, but unfortunately, such information is rarely available. 

The amount of organic material in a soil is the product of a number of di- 
verse factors (Jones, 1973) acting over a period of time on the relative rates 
of the return of organic residues to the soil and their subsequent breakdown 
in the soil. In “savanna”, Jones considers, it is probably the effective length 
of the wet season that is more important in determining soil organic matter 
levels than annual rainfall. Birch & Friend (1956) recognised one major fac- 
tor, namely rainfall, as having an overall influence on soil organic matter 
status in East African soils and two minor factors, temperature and clay con- 
tent, that exercise a modifying influence. 

Conclusion 

In writing this review, some speculation has been included because of a 
lack of sufficient data relevant to tropical conditions. It is hoped that this 
paper will, to some degree, provide an impetus for further work. There is 
obviously a great need for comparative data on the fates of primary produc- 
tion in tropical and subtropical biomes of various types, including detail 
within the dead matter consumption and decompostion components. Fur- 


76 


Journal of South African Botany 


ther study of soil organic matter levels would also appear to be appropriate 
as conflicting accounts abound in the literature. 

Acknowledgements 

The contributions made to this review by E. H. Bucher, P. R. Furniss, 
M. Lamotte, R. G. Noble and B. Pendle are acknowledged. I thank Jane 
Theaker, J. Bezuidenhout, P. Ferrar, J. L. Flutson, E. Madsen, J. C. Me- 
naut and anonymous reviewers for commenting on the manuscript. 


References 

Allison, F. E., 1973. Soil organic matter and its role in crop production. Amster- 
dam: Elsevier. 637 p. 

Anderson, J. M. and Macfadyen, A. (eds.), 1976. The role of terrestrial and aquatic 
organisms in decomposition processes. Oxford: Blackwell Scientific Publica- 
tions. 474 p. 

Anon., 1978. Organic matter in tropical soils (1947-1966) and Organic matter in trop- 
ical soils (1965-1976). Harpenden: Commonwealth Bureau of Soils. 19 p. 
Anon., 1979a. Secondary production: consumption and decomposition. In: Anon., 
Tropical grazing land ecosystems. Paris: UNESCO/UNEP/FAO. pp. 

146- 206. 

Anon., 1979b. Soils and erosion. In: Anon., Tropical grazing land ecosystems. Paris: 
UNESCO/UNEP/FAO. pp. 96-118. 

Ashton, D. H., 1975. Studies of litter in Eucalyptus regnans forests. Aust. J. Bot. 
23: 413-433. 

Bille, J. C., 1978. Etude de la production primaire nette d'un ecosysteme sahelien. 
Paris: These Universite Paris Sud. 82 p. 

Birch, H. F. and Friend, M. T., 1956. The organic matter and nitrogen status of 
East African soils. J. Soil Sci. 7: 156-67. 

Bornemissza, G. F., 1979. The Australian Dung Beetle Research Unit in Pretoria. 
5. Afr. J. Sci. 75: 257-260. 

Brady, N. C., 1974. The nature and properties of soils. New York: Macmillan. 639 p. 
Bucher, E. EL, 1982. Chaco and caatinga — South American arid savannas, wood- 
lands and thickets. In: Huntley, B. J., and Walker, B. H. (eds.), The ecol- 
ogy of tropical savannas Berlin: Springer-Verlag. pp. 48-79. 

Ferrar, P., 1982a. Termites of a South African savanna. I. List of species and sub- 
habitat preferences. Oecologia (Berl.) 52: 125-132. 

Ferrar, P., 1982b. Termites of a South African savanna. IV. Subterranean popu- 
lations, mass determinations and biomass estimations. Oecologia (Berl.) 52: 

147- 151. 

Greenland, D. J. and Nye, P. H., 1959. Increases in the carbon and nitrogen con- 
tents of tropical soils under natural fallows. J. Soil Sci. 10: 284-299. 
Greenland, D. J. and Kowal, J. M. L., 1960. Nutrient content of moist tropical 
forest of Ghana. PI. Soil 12: 154-174. 

Harris, W. V., 1955. Termites and the soil. In: Kevan, D. K. M. (ed.). Soil Zool- 
ogy. New York: Academic Press, pp. 62-72. 


Decomposition and reduction of soil organic matter in African biomes 77 

Huntley, B. J. and Morris, J. W. 1982. Nylsvley— the structure of a South African 
savanna. In Huntley, B. J. and Walker, B. H. (eds.), The ecology of tropi- 
cal savannas. Berlin: Springer-Verlag. pp. 433-445. 

Jenny, H., 1950. Causes of the high nitrogen and organic matter content of certain 
tropical forest soils. Soil Sci. 69: 63-69. 

Jones, M. J., 1973. The organic matter content of the savanna soils of West Africa 
/. Soil Sci. 24: 42-53. 

Kononova, M. M., 1961. Soil organic matter its nature, its role in soil formation and 
in soil fertility. Oxford: Pergamon Press. 450 pp. 

Lamotte, M., 1975. The structure and function of a tropical savanna ecosystem. In: 
Golley, F. B. and Medina, E. (eds.). Tropical Ecological Systems. Trends in 
terrestrial and aquatic research. Ecological Studies 11. Berlin: Springer-Ver- 
lag. pp. 179-222. 

Lamotte, M., 1977. Observations preliminaries sur les flux d’energie dans un eco- 
systeme herbace tropical, la savanne de Lamto (Cote d’Ivoire). Geo-Eco- 
Trop 1: 45-63. 

Lamotte, M., 1978. La savane preforestiere de Lamto, Cote d’Ivoire. In: Lamotte, 
M. and Bourliere, F. (eds.), Problemes d’ecologie: Structure et functionne- 
ment des ecosystemes terrestres. Paris: Masson, pp. 231-311. 

Lamotte, M., 1979. Structure and functioning of the savanna ecosystems of Lamto 
(Ivory Coast). In: Anon., Tropical grazing land ecosystems. Paris: UNES- 
CO/UNEP/FAO . pp. 511-561. 

Lee, K. E. and Wood, T. G., 1971. Termites and soils. London: Academic Press. 
251 p. 

Malaisse, F., 1978. High termitaria. In: Werger, M. J. A. (ed.), Biogeography and 
ecology of Southern Africa. The Hague: Junk. pp. 1297-1300. 

Malaisse, F. and Anastassiou-Socquet, F., 1977. Phytogeographie des houtes 
termiteres du Shaba meridional (Zaire). Bull. Soc. r. Bot. Belg. 110 : 85-95. 

Malaisse, F., Freson, R., Geoffinet, G. and Malaisse-Mousset, M., 1975. Litter 
fall and litter breakdown in Miombo. In: Golley, F. B. and Medina, E. 
(eds.), Tropical Ecological Systems. Trends in terrestrial and aquatic re- 
search. Ecological Studies 11. Berlin: Springer-Verlag. pp. 137-152. 

Mohr, E. C. J., Van Baren, F. A. and Van Schuylenborgh, J., 1972. Tropical 
soils. The Hague: Mouton-Ichtiar Baru-Van Hoeve. 480 p. 

Morris, J. W., Bezuidenhout, J. J. and Furniss, P. R., 1982. Litter decomposition 
at Nylsvley. In: Huntley, B. J., and Walker, B. H. (eds.), The ecology of 
tropical savannas. Berlin: Springer-Verlag. pp. 535-554. 

Nye, P. H. and Greenland, D. J., 1960. The soil under shifting cultivation. Techni- 
cal Communication 51. Harpenden: Commonwealth Bureau of Soils. 156 p. 

Olson, J. S., 1963. Energy storage and the balance of producers and decomposers in 
ecological systems. Ecology 44 : 322-331. 

Petal, J., 1978. The role of ants in ecosystems. In: Brian, M. V. (ed.), Production 
ecology of ants and termites. Cambridge: Cambridge University Press, pp. 
293-325. 

Rodin, L. E. and Bazilevich, N. I., 1967. Production and mineral cycling in terres- 
trial vegetation. Edinburgh: Oliver & Boyd. 

Russell, E. J., 1961. Soil conditions and plant growth. 9th ed. London: Longmans 
688 p. 


78 


Journal of South African Botany 


Sanchez, P. A., 1976. Properties and management of soils in the tropics. New York: 
Wiley. 412 p. 

Schnitzer, M. and Khan, S. U., 1972. Humic substances in the environment. New 
York: Marcel Dekker. 

Singh, J. S. and Gupta, S. R., 1977. Plant decomposition and soil respiration in ter- 
restrial ecosystems. Bot. Rev. 43: 449-528. 

Trapnell, C. G., Friend, M. T., Chamberlain, G. T. and Birch, H. F., 1976. The 
effects of fire and termites on a Zambian woodland soil. J. Ecol. 64 : 
577-588. 

Vine, Ft., 1968. Developments in the study of soils and shifting agriculture in tropical 
Africa. In: Moss, R. P. (ed.), The soil resources of tropical Africa. Cam- 
bridge: Cambridge University Press, pp. 89-119. 

Wood, T. G., 1976. The role of termites (Isoptera) in decomposition processes. In: 
Anderson, J. M. and Macfadyen, A. (eds.), The role of terrestrial and aqua- 
tic organisms in decomposition processes. Oxford: Blackwell Scientific Pub- 
lications. pp. 145-168. 

Wood. T. G. and Sands, W. A., 1978. The role of termites in ecosystems. In: Brian, 
M. V. (ed.), Production ecology of ants and termites. Cambridge: Cam- 
bridge University Press, pp. 245-292. 


J1 S. Afr. Bot. 49 (1): 79-86 (1983) 


Book Reviews 

Progress in Botany 43: Morphology, Physiology, Genetics, Taxonomy, 
Geobotany, edited by H. Ellenberg, K. Esser, K. Kubitzki, E. Schnepf and 
H. Ziegler, with pp. xii + 382. Berlin, Heidelberg, New York: Springer- 
Verlag, 1981. DM 129,00. ISBN 3-540-11091-7. 

Progress in Botany, or by its alternative title, Fortschritte der Botanik, is not as 
widely known in South Africa as it should be. In these times of extreme specialisa- 
tion, and fashions, it is refreshing to find such a collection of reviews which encom- 
passes so many fields of botanical enquiry. Its special usefulness in South Africa is 
that it draws attention to a very broad spectrum of literature, but particularly to the 
views of Continental botanists, greatly facilitated by all but two of the chapters being 
in English. 

Some reviews are very condensed, occasionally being little more than notes, 
whilst the best are very readable summaries spanning the work of a number of years. 
The references are extensive and up to date. A valuable feature is that many chap- 
ters are prefaced by notification of comprehensive new books, major surveys and 
bibliographies. One of the attractions of this series has been its coverage of less fam- 
iliar topics, such as the chapter “Special Cytology” by E. Schnepf and the valuable 
guide to modern ideas in floral morphology by P. K. Endress. Two other contribu- 
tions are worthy of special notice; thus Stephan Vogel gives a most readable review 
of work in floral biology since 1979, and Clive Jensen discusses proteins in plant evol- 
ution and systematics. A further major and valuable contribution is the wide-ranging 
review of the systematics and evolution of higher plants by Poppendieck, related in a 
stimulating and humorous style. 

Geobotany, or perhaps more explicitly, vegetation science, is represented by 
Knapp's survey of concepts and regions, from which however Africa is noticeably ab- 
sent, and, more specialised, a history of quarternary floras of E. Europe and Central 
Asia by Frenzel. A well written chapter of wider appeal is that by W. Schmidt, in 
which he provides an overview of productivity and ecosystem research, including lit- 
ter and nutrient cycling, and terrestrial and aquatic biomass productivity. 

The plant physiology section includes a most informative and comprehensive es- 
say by Amesz on the photosynthetic reaction centre. Ideas on the metabolism of or- 
ganic acids are examined by Kluge, and of inorganic nitrogenous compounds by 
Kessler, and secondary products by Schiitte. The section is concluded by a short re- 
view of recent papers on the metabolism and mode of action of gibberellins, brassi- 
nolide and cytokinins. The excellent genetics section is of much wider scope than is 
usual in purely botanical reviews. Nagel deals with aspects of DNA replication, 
Binding and Nehls with recombination in higher plants, including protoplast fusion, 
and Gottschalk with mutation and mutagenesis. The genetics of storage protein an 
regulation of gene expression is reviewed by Blaich, and Hagemann and Borner give 
a useful account of advances in plastid inheritance. 

Many of the articles, although giving extensive references, are not very cri ica . 
Robard’s summary of the present state of knowledge of plastid structure an in er 
relationship can only be described as terse, and Wohrmann and Tomiu^ ar y 
justice to population genetics in a mere ten pages. Whether by intention or no , 


79 


80 


Journal of South African Botany 


coverage of non-flowering plants with the exception of the fossil forms discussed by 
F. Schaarschmidt, is scanty. A more serious criticism is the inadequacy of the index: 
there are too few comprehensive headings, and thus the inclusion of minor interest 
items, to the detriment of the more probable reader interest, although one must ad- 
mit that a fairly detailed contents list does to some extent compensate for this. Cross- 
referencing is very incomplete: thus Umbelliferae and Apiaceae, and Gramineae and 
Poaceae each refer to separate entries, but not to any in common. More slipshod is 
to allow “legumes” to stand in the text and index for the family name, and needless 
to say there are unrelated page references to Fabaceae and to each of the tribes of 
the Leguminosae! Editorial shortcomings (or is it editorial superfluity?) also lead in 
places to language awkwardness, and to a number of spelling errors. 

Progress in Botany is published annually. The book is offset from typescript but 
the layout is clear and not unpleasing. There is a strong and attractive binding and by 
current standards the price should not be regarded as excessive. This is a valuable 
source of reference and the acquisition of the series can be emphatically recommend- 
ed to all botanical libraries catering for undergraduate and research interests. 

A. R. A. Noel 


Biology of Plants, by P. H. Raven, R. F. Evert and H. Curtis, with pp. xvi 
+ 686, 3rd edition. New York: Worth Publishers Inc., 1981. US $19,95. 
ISBN 0-87901-132-7. 

Botany: An Introduction to Plant Biology by Weier et al. (Wiley), the 6th edition 
of which appeared this year, has for the past two decades been one of the best and 
one of the most popular textbooks of plant biology. However, it most certainly has 
found a rival in the 3rd edition of Biology of Plants, whose 2nd edition predecessor 
made a successful debut when R. F. Evert of the University of Wisconsin joined as 
second author. 

I would agree with reviewer R. A. Quatrano (Nature 295: 471, 1982) that this 
text must be “considered as one of the best overall books of general botany.” From 
cover to cover — in fact including the cover which features a reproduction of Vincent 
van Gogh’s Irises — it becomes abundantly clear that considerable teaching and pro- 
fessional experience has gone into a skilfully planned and comprehensively written 
text. Not only have the personal experiences of Raven and Evert themselves been 
brought to bear, but also those of many of their colleagues at a number of U.S. uni- 
versities and even as far afield as Australia, Great Britain, the Federal Republic of 
Germany and South Africa. About 10 botanists assisted with material, discussions, 
guidance, and revisions of certain chapters, and nearly 50 others supplied informa- 
tion of various kinds. 

Few textbooks can be as well illustrated, either photographically or diagrammati- 
cally, and extending colour photomicrography to angiosperm anatomy will ensure 
that the internal architecture of the plant is viewed afresh. While referring to anat- 
omy, it is fair to point out that Prof. Evert’s hand is clearly evident in the presenta- 
tion of the plant cell, the angiosperm plant body’s structure and development, and 
the morphology generally. This also applies to the detail and superb quality of his 
light and electron micrographs. 


Book Reviews 


81 


In all there are eight major sections with a total of 30 chapters, four appendices 
(chemistry, metrication, classification of organisms, geologic eras), a 26 page double- 
column glossary, and a 17-page four-column subject index. An interesting feature is 
the essay , a factual review of a topical, novel or perplexing phenomenon, written 
succinctly and italicized for ready identification. Only nine of the 30 chapters have no 
special essay, but then some have two or more, making a total of 47 (as compared to 
33 in the 2nd edition). Chapter 11 on the Procaryotes includes no less than 7 essays 
on subjects such as genetic engineering, mycoplasmas and viroids, to mention only 
three. Each chapter, furthermore, is provided with a summary and in most cases also 
with suggestions for further reading. 

When compared to the 2nd edition the new book features a number of improve- 
ments, one of which is updated information. Examples: photosynthesis, plant growth 
regulators, and uptake and transport. Another is the inclusion of information relat- 
ing to the rapid advancement of biology during the latter part of the 1970’s. Ex- 
amples: recombinant DNA, tumour-inducing plasmids, and lectins. Further improve- 
ments concern reorganisation and emphasis. Examples: the chemistry of heredity has 
been transferred from the plant cell section to that of genetics and evolution; under 
the section on diversity, the blue-green algae are treated as the cyanobacteria; slime 
and water moulds are included in an own chapter on heterotrophic protista; emphasis 
on fossil plants and progymnosperms has been increased. 

How might this book serve as a teaching text? This will of course depend on the 
university system in question. That is, is the curriculum based on a quarter, semester 
or annual system? Is the course at the introductory, prerequisite, undergraduate or 
graduate level? Are the students botany or biology majors, or is botany a non-major 
subject? Is it intended as preparatory to qualifying examinations? The text would be 
an excellent choice in the traditional three-year, nine-course, double-major degree or 
the ten-course, single-major degree, where it could be adopted unchanged for 
Botany 1. (A practical manual and preparation compendium* are also available). 

It is a good text for a semester course although here, and more particularly in a 
quarter course, specific chapters would have to be selected to fit the teaching sched- 
ule. However, this should not present problems as the book is so organized that top- 
ics can be assigned in any preferred combination of sequences. The text may be less 
well suited to a system (such as in Sweden) where undergraduate courses are numer- 
ous but of short duration. However, even in this case the book offers the student 
what the system often fails to provide, namely the perception, synthesis and inte- 
gration of the subject. 

It becomes not so much a question as to what level the book should be aimed, 
but rather one of whether it makes accessible the basic concepts of botany as a dy- 
namic subject. With its depth, which is sufficient for an introductory text, and its en- 
cyclopedic breadth of subject matter on which more advanced courses in physiology, 
morphology and ecology could build, Biology of Plants is the kind of text that the 
student can use even without the guidance of a watchful tutor. For only US $20,00 it 
is excellent value and deserves to be widely adopted. 


Chris H. Bornman 


* Laboratory Topics in Botany, by R. F. Evert and S. E. Eichhorn, with pp. vii + 
176. New York: Worth Publishers, Inc., 1981. ISBN 0-87901-142-A 
Preparation Guide for Laboratory Topics in Botany, by S. E. Eichhorn, J. w. 
Perry and R. F. Evert, with pp. vi + 168. New York: Worth Publishers, Inc., 
1981. ISBN 0-87901-061-4. 




82 Journal of South African Botany 

Plant Growth Regulators: Agricultural Uses, by L. G. Nickell, with 
pp. xii + 173, 29 figures. Berlin, Heidelberg, New York: Springer-Verlag, 
1982. DM 47, 50, approx. US $22,10. ISBN 3-540-10973-0. 

This neatly and well produced book was not written in an attempt to present a 
comprehensive treatise of all the agricultural uses, or potential uses of growth sub- 
stances. The author himself stresses in the preface that little attention is given to the 
basic research which formed the backbone for the successful exploitation of these 
compounds. Emphasis was rather placed on the positive results which have already 
been achieved by applying the obtained data in the various fields of agriculture. By 
setting these terms of references the author succeeded in producing a well laid-out 
and easy to read text which clearly indicates the tremendous advancements made in 
the use of plant growth regulators in agriculture over the last decades. Throughout 
the book the author has succeeded in maintaining a good balance between what has 
been achieved and what could reasonably be expected in this field in future. 

The data presented is covered in 24 chapters. This may seem an unnecessarily 
large number but these were necessitated by the fact that the author chose to discuss 
the available information by plant processes which are controlled or affected by the 
plant hormones. This approach makes the book easy to read and both undergrad- 
uates and graduate students should benefit from it. The large number of references 
to which the author refers should prove of great benefit not only to students but also 
to intending or existing researchers in this field. For these reasons all teaching and 
research institutions should have this book on its shelves. 

J. van Staden 


Silicon and Siliceous Structures in Biological Systems, edited by T. L. 
Simpson and B. E. Volcani, with pp. xvi + 587, 515 figures and 15 tables. 
Berlin, Heidelberg, New York: Springer-Verlag, 1981. DM 240, approx. US 
$111,80. ISBN 0-387-90592-8. 


Siliceous structures occur in many organisms in which amorphous hydrated silica 
forms cell walls, scales, tests and other skeletal features. In some groups, e.g. the 
diatoms, the morphology of these siliceous structures form the basis for their tax- 
onomy. Silicon is also required for numerous metabolic processes and is essential for 
growth and bone development. It is, however, still not clear in what soluble form the 
silicon is transported and what determines the morphological pattern of siliceous 
structures. 

An international symposium titled, “Siliceous structures and silicon deposition in 
living organisms”, which was held in Richmond, Virginia, U.S.A., during December 
1978, initiated the publication of this book. 

In the preface to the book the editors state that its publication was undertaken 
with the dual purpose of bringing together information on the deposition by living 
organisms of unique skeletal structures composed of amorphous silica, and reviewing 
recent data on the involvement of silicon in physiological processes. The book also 
discusses the ultrastructural, physiological and biological aspects of silica. 

Attention is focused upon the biological aspects of silica and siliceous structures, 
with the emphasis on the evolution, phylogeny, morphology and distribution of these 
structures, on the cellular aspects of silica deposition and on the physiological and 


Book Reviews 


83 


biochemical roles of silicon. This is the first time that this wealth of data has been 
compiled into one volume. The field covered is therefore extensive and accordingly it 
is recommended as a useful reference work for workers in such diverse fields as bio- 
chemistry, botany, cell biology, diatomology, marine biology, micropaleontology, 
phycology, protozoology and zoology. 

This is a most informative book and makes stimulating reading. The various 
chapters are all excellent reviews of the subjects in question. Each chapter is rich in 
references including the most recent (1980) at the time of going to press. Although 
the book covers a wide field, the editors should be complimented on their efforts to 
cross-reference certain features, structures, etc., relevant to the various chapters. 

The book comprises 18 chapters written by experts in their relevant fields. In the 
introduction the editors present a brief survey of the chapters, plus notes and 
thoughts on morphogenesis of silica in biological systems and the evolution of silica 
deposition. The bulk of the book is divided into two sections: 

1. Physiological aspects of silicon. Three chapters. Silica in cellular metabolism of di- 
atoms, germanium-silica interactions in biological systems, and silicon in bone forma- 
tion. 

2. Siliceous structures: distribution, deposition, ultrastructure and morphogenesis. 
Fourteen chapters dealing with algal groups, protozoa, sponges and higher plants. 

The general layout of the chapters is good with notes (where applicable) on ter- 
minology and procedures. The discussions and the posing of questions make it stimu- 
lating reading. There is some overlap between chapters 6 and 7 regarding the charac- 
teristics of the diatom cell. Although not detracting from the importance and 
usefulness of this book it is a pity that throughout chapter 7 the identity of the 
diatom referred to as Navicula pelliculosa is incorrect. The true identity of N. pelli- 
culosa has been established in various European publications. 

The book is excellently produced with only some minor spelling and printing 
errors. The bibliography is bounteous with just on 1 000 items in total and number- 
ing between 23 and 108 items per chapter. In some of the chapters there appears to 
be some bias in favour of English language references. The chapters are well illus- 
trated with figures of high quality. Figure captions are usually adequate and accurate 
but occasionally fail to give the method of photography employed. A most useful 36- 
page subject index, which appears to be fairly comprehensive, completes the book. 

This is a highly recommended reference work for both under- and post-graduate 
study and should be acquired by all biological reference libraries. Unfortunately, the 
high price may deter individuals and specialists interested in only a section of the 
field covered, from purchasing this book. 


F. R. Schoeman 


Physiological Plant Ecology 1: Responses to the Physical Environ- 
ment, edited by O. L. Lange, P. S. Nobel, C. B. Osmond and H. Ziegler, 
with pp. xv + 625, 110 figures. Berlin, Heidelberg, New York: Springer- 
Verlag, 1981. Volume 12 in “Encyclopedia of Plant Physiology, New 
Series”. DM 239, approx. US $111,30. ISBN 3-540-10763-0. 


It is high time that plant biologists sorted out the difference between physio ogi 
cal plant ecology and plant ecophysiology. The former emphasises ecology, t e a er 
physiology. There is obviously an area of overlap but, at least in my view, p ysio ogy 


84 


Journal of South African Botany 


is aimed at what goes on inside a plant and ecology deals with the levels of inte- 
gration above that. Failure to distinguish between the two has detracted from the 
value of this first part of the four volume series, since, although there are a number 
of interesting and informative chapters, the book is in general neither one nor the 
other. It is noteworthy that although the book is titled Physiological Plant Ecology , 
the editors have titled their own introductory chapter “Perspectives in Ecological 
Plant Physiology”. This review is written from the perspective of an ecologist. 

This first part of the four-part series, Responses to the Physical Environment , 
consists of 17 chapters by 22 contributors (in addition to an Introduction by the edi- 
tors), making up 625 pages of well-printed and easily readable type. It has been well 
produced with very few errors. There are seven chapters on radiation and responses 
to radiation in terrestrial plants, two on the aquatic environment (one of which is 
specifically on light). Five chapters deal with temperature, and there is one each on 
wind, fire and the soil environment. 

The best chapters are those which deal in some depth with the physiological re- 
sponses of plants to various aspects of the environment. It is a pity that the editors 
did not try to achieve a more consistent framework for the chapters. 

The sections on Ecological Implications, at the ends of chapters 5 and 6, were 
useful contributions and would have enhanced the value of the other chapters. 

Chapter 8, on the aquatic environment, is too superficial to be of much use, as 
are the final three chapters on wind, fire and the soil environment. Each of these 
four topics requires a book of their own to do them any justice. Chapter 11, on 
micro-organisms, is out of place. 

In general, the first five chapters on light and chapters 10, 12, 13 and 14 on tem- 
perature are good contributions and make the book worth-while. 

The bibliographies are up to date and most of them are comprehensive. There is 
both an author and subject index, and the latter is well constructed. The level of the 
book is suited mainly for final year undergraduate and postgraduate students. How- 
ever, at US $112 it is destined mainly for libraries, where it will be a useful refer- 
ence. 


B. H. Walker 


Hormonal Regulation of Development I, edited by J. Macmillan, with pp. 
vi + 681. Berlin, Heidelberg, New York: Springer-Verlag, 1980. Volume 9 
in “Encyclopedia of Plant Physiology, New Series”, edited by A. Pirson and 
M. H. Zimmerman. US $134,60. ISBN 3-540-10161-6. 

This is the first of a set of three volumes in the New Series of the Encyclopedia of 
Plant Physiology intended to cover the field of plant growth substance-controlled 
regulation of growth and development. The planning of the three volumes was the 
responsibility of N. P. Kefford of the University of Hawaii. In Volume I consider- 
ation is given to the molecular and subcellular components of hormonal regulation, 
and a refreshing approach has been adopted. Instead of treating each class of hor- 
mone separately, an attempt is made to examine the properties and principles com- 
mon to all classes. Kefford argues that such an approach allows examination of the 
hypothesis that differing classes of hormones, “acting according to common prin- 
ciples, are determinants of processes and phases in plant development”. This seems a 
logical approach since it is assumed that the major groups of plant hormones (and 


Book Reviews 


85 


other growth- affecting substances) act separately, sequentially or in concert on the 
basic processes of plant growth and development. In the Introduction, the editor, J. 
MacMillan of the University of Bristol, presents the scope of the volume and gives a 
chapter synopsis in a way that makes much additional review comment seem redun- 
dant. 

There are six chapters. Chapter 1 gives a brief history of the five classes of hor- 
mones, fists the individual substances, and their occurrence, and includes other con- 
stituents which affect plant growth. Chapter 2 deals with extraction, purification, and 
identification techniques, and presents some case histories. No mention, however, is 
made of ethylene, not even under the heading Identification without isolation. In 
chapter 3 quantitative analytical aspects are dealt with. This is a provocative chapter 
seeking, as pointed out by the editor, to stimulate critical appraisal of the accuracy of 
methods used by investigators. Again, reference to ethylene is missing. Immunolog- 
ical assays are mentioned briefly, although their increased use and eventual superces- 
sion of the bioassay is foreseen. (Developments in 1980 and 1981 subsequently 
showed that radioimmunoassay of indoleacetic acid, abscisic acid and other hor- 
mones has become a reality.) 

Chapter 4 is the heart of the book, not only in its length but in its very thorough, 
scholarly treatment of the biosynthesis and metabolism of all five hormone classes. 
Chapters 5 and 6 review receptor sites and molecular effects, respectively. Finally, 
the book contains nearly 100 pages of author and subject indexes. 

This is the kind of reference work that sells itself, even though it has been very 
well written, edited and produced. It probably was a major task to limit its length 
(681 pages compared to 1 357 of its counterpart 21 years earlier), but this obviously 
is a factor that has kept the costs down. The price of US $134.60 (or $107.70 for pur- 
chasers of Vols 9, 10 & 11) will ensure that this book finds its way also to the private 
bookshelf. 


Chris H. Bornman 


Plant Growth Substances 1979, edited by F. Skoog, with pp. vi + 527 and 
209 figures, Berlin, Heidelberg, New York: Springer-Verlag, 1980. US 
$57.90. ISBN 3-540-10182-9. 

This volume represents the proceedings of the 10th International Conference on 
Plant Growth Substances held in Madison, Wisconsin in July 1979. As a proceeding s 
volume it is in a class of its own and one that editors of past and especially future 
conference proceedings would be wise to consult. . 

The publication has the appearance of a text book: there is uniformity ot sty e 
and ample evidence that much thought and careful planning had gone into its prep- 
aration. The conference itself was organised into 11 sections, each with a chairman 
who was responsible for the arrangement of the topic material and the invite re 
ports. These chairmen not only introduced the topics and led the discussions, u 
also assisted in the editing of the invited reports. ... , 

Each of the five classes of plant growth substances (auxins, cytokimns, gi er 
fins, abscisic acid and ethylene) is represented in a section. Four other close y re a e 
sections are devoted to new growth factors (this comprises a two-page summary), 
hormonal regulation in plant reproductive development, hormonal regu a ion i 
morphogenesis, and agricultural uses of plant growth regulators. The remaining 


86 


Journal of South African Botany 


sections include a historic review and a commemorative symposium. The former, 
under the title Origin and development of plant growth substance research, includes 
reviews respectively by J. Heslop-Harrison and K. V. Thimann on Darwin and the 
movement of plants and the development of plant hormone research since f920. The 
latter section on plant movements was organised and presided over by A. W. Gal- 
ston to commemorate the publication in 1880 of Darwin’s The Power of Movement in 
Plants. 

A total of 244 abstracts were submitted to the conference, but only 50 reports 
were accommodated in the proceedings, indicating that a strict selection procedure 
was applied. The papers reflect this. In the sections on the individual classes of hor- 
mones most of them report current states of knowledge, many dealing with meta- 
bolic aspects, and a number with mechanisms of action and physiological roles in 
plants. In a departure from previous proceedings, the rapidly expanding practical 
utilisation of plant hormone research is acknowledged in separate sections on mor- 
phogenesis, reproduction and agricultural application. The section on agricultural 
uses of plant growth substances is one of the most interesting, for it is here that the 
expanding and galloping applications of gibberellin and ethylene, respectively, are 
considered. Within its manifold effects ethylene’s influence spans all phases of plant 
development. As D. R. Dilley puts it: “Ethylene, indeed, is a busy gas.” 

As was the case with previous Plant Growth Substance Conference proceedings, 
that of the 10th meeting will be an essential aid to the plant hormone physiologist. 

Incidentally, the 11th conference took place in Aberystwyth in July 1982. 


Chris H. Bornman 


J1 S. Afr. Bot. 49 (2): 87-101 (1983) 


THE DEFINITION OF LEAF CONSISTENCE CATEGORIES IN THE 
FYNBOS BIOME AND THEIR DISTRIBUTION ALONG AN 
ALTITUDINAL GRADIENT IN THE SOUTH EASTERN CAPE 

R. M. Cowling 

(Department of Botany, University of Cape Town, Private Bag, Rondebosch 
7700, R.S.A.) 

B. M. Campbell* 

( Botanical Research Unit, P.O. Box 471, Stellenbosch 7600, R.S.A.) 
Abstract 

Leaf consistence is a morphological feature which is used as an attribute for the 
structural classification of communities and for the evolutionary interpretation of 
plant form and function. There are, however, no unambiguous criteria for the 
characterisation of leaf consistence categories. In this paper criteria for the subjec- 
tive categorisation of leaf consistence by “feel” are tested by determining indexes of 
sclerophylly and succulence on leaves classed on a a priori basis as sclerophyll, ortho- 
phyll, fleshy (semi-succulent) or succulent. Results show differences in degree of 
sclerophylly (leaf dry weight per unit leaf area) and degree of succulence (maximal 
water content per unit leaf area) among the subjectively determined categories. The 
adaptive significance of the leaf consistence categories is indicated by direct gradient 
analysis of two parallel transects in the south eastern Cape, one in fynbos, the other 
in non-fynbos vegetation. In non-fynbos changes in leaf consistence along the gra- 
dient are interpreted largely as a response to changes in climate and soil moisture. 
We interpret the predominance of sclerophyll leaves throughout the fynbos gradient 
as a response to low soil fertility. 

UlTTREKSEL 

DEFINERING VAN BLAARSTRUKTUUR-KATEGORIEE IN DIE FYNBOS 
BIOOM EN DIE VERSPREIDING DAARVAN LANGS 'N HOOGTE GRA- 
DIENT IN DIE SUIDOOS-KAAP 

Blaarstruktuur as ’n morfologiese kenmerk word gebruik vir die klassifikasie van 
gemeenskappe en die evolusionere interpretasie van plantvorm en funksie. Daar be- 
staan egter geen duidelike kriteria vir die subjektiewe indeling van blaarstruktuurka- 
tegoriee nie. In hierdie artikel word die subjektiewe indeling van blaarstruktuur deur 
die “voel” metode getoets deur indekse van sklerofillie en sukkulensie van blare op 
’n a priori basis as sklerofil, orthofil, vlesig of sukkulent aangegee. Resultate toon n 
verskil in die graad van sklerofillie (blaar droe gewig per eenheid blaar oppervlakte) 
en graad van sukkulensie (maksimale water inhoud per eenheid blaar oppervlakte) 
onder die subjektief bepaalde kategoriee. Die aanpasbaarheidsbetekenis van blaar- 
struktuur-kategoriee is aangedui deur ’n direkte gradient ontleding van twee parallel- 


* Present address: University of Zimbabwe, Harare, Zimbabwe. 
Accepted for publication 12th November, 1982 


87 


88 


Journal of South African Botany 


le transekte in die Suidoos-Kaap, een in fynbos en die ander in nie-fynbos plante- 
groei. In nie-fynbos kan die veranderings interpreteer word as ’n reaksie op klimaats- 
en grondvog. Ons interpreteer die oorwegende sklerofille blaar-voorkoms in die fyn- 
bos gradient as 'n reaksie op die lae grondvrugbaarheid. 

Introduction 

Consistence is here defined to refer to both the firmness and thickness of 
a leaf (see Werger and Ellenbroek, 1978). This leaf characteristic has long 
been cited as a structural feature which correlates well with the regional cli- 
matic regime (e.g. Schimper, 1903). The adaptive significance of leaf con- 
sistence has been the focus of a number of studies (Loveless, 1962; Beadle, 
1966; 1968; Small, 1972; 1973; Orians and Solbrig, 1977; Werger and Ellen- 
broek, 1978; Camerik and Werger, 1980; Bond, 1981), while leaf consist- 
ence per se is used as an attribute for structural classification with purely 
phytosociological objectives (Webb, Tracey, Williams and Lance, 1970; 
Bond, 1981). The definition of leaf consistence categories, however, and 
their recognition in the field, remains problematic (Loveless, 1962; Small, 
1973; Beadle, 1966; Bond, 1981). 

Leaf consistence grades from an orthophyll or soft-leaved type towards 
increasing sclerophylly. All categories are subjective and cannot be sharply 
defined by any standards (Beadle, 1966). Students of vegetation in the Fyn- 
bos Biome (Kruger, 1978) have achieved some measure of agreement on a 
subjective classification of leaf consistence types based on “feel” (Bond, 
1981: Table 1). Here we provide a test for the subjective criteria used in this 
scheme (see Table 1) by determining indexes of sclerophylly and succulence 
(Camerik and Werger, 1980) on leaves, categorised on an a priori basis as 
orthophyll, sclerophyll, fleshy (semi-succulent) or succulent. 

Direct gradient analysis (Whittaker, 1967) has proved to be an effective 
tool in highlighting the relations of plant community structure to environ- 
ment (Mooney and Harrison, 1972; Mooney, Gulmon and Parsons, 1974; 
Parsons and Moldenke, 1975; Werger and Ellenbroek, 1978; Cowling and 
Campbell, 1980; Bond, 1981). We briefly indicate the adaptive significance 
of leaf consistence by examining the relative importance of consistence cat- 
egories along parallel altitudinal gradients in fynbos and non-fynbos vegeta- 
tion in the south eastern Cape. 

Material and Methods 
Consistence determination 

Branches were sampled from a range of species subjectively classed into 
consistence categories (Table 4) by ourselves in the field. All samples were 
collected at llhOO in mid-December from species growing in the National 
Botanic Gardens, Kirstenbosch. Soil type and microclimate were regarded 


Leaf consistence categories in Fynbos Biome and distribution 


89 


Table 1. 

Subjective categorisation of leaf consistence by “feel” [adapted from Bond (1981)]. 


Texture 

Characterisation and examples 

Succulent 

Leaves turgid, express copious liquid when 
squeezed between thumb and forefinger Delosper- 
ma, Mesembryanthemum, Crassula, Euphorbia 
mauritanica 

Fleshy (Semi-succulent) 

Leaves usually pulpy, little liquid expressed but 
leaves collapse with a rubbery or gelatinous texture 
when rolled between thumb and forefinger Relha- 
nia genistaefolia, Chrysocoma tenuifolia, Pteronia 
fas data, Felicia flifolia 

Sclerophyll 

Leaves hard, coriaceous and thick, breaking when 
folded Protea, Phylica, Passerina, Sideroxylon in- 
erme 

Orthophyll 

Leaves soft, thin and pliant when folded Ocotea 
bullata, Clutia pulchella, Gnidia coriacea, lndigofe- 
ra denudata 


as uniform for the whole sample. No visible signs of leaf chlorosis or flaccidi- 
ty were noted. 

Sampled branches were sealed in plastic bags and immediately trans- 
ported to the laboratory. Ten leaves were selected from each species and 
submerged in a container of distilled water and kept in a dark coldroom 
(10 °C) for 24 h to achieve maximum hydration. The next day fresh mass 
and the surface area of each leaf were determined. Dry mass was determin- 
ed by oven-drying the leaves at 105 °C for 24 h. With these data the follow- 
ing parameters could be calculated (Camerik and Werger, 1980): 

Degree of sclerophylly, defined as the leaf dry mass per unit (bifacial) leaf 
area (g dm -2 ). 

Degree of succulence defined as the maximal water content of the leaf per 
unit (bifacial) leaf area (g dm -2 ) 

We did not use Loveless’ (1962) index of sclerophylly (per cent crude fibre 
plus per cent crude protein) as it would be of doubtful value in distinguish- 
ing between orthophyll and succulent leaves. 

Direct gradient analysis 

The vegetation transects were located in the Gamtoos river valley (S33 
50', E24° 55'), from near the village of Hankey on the valley floor to near 


Table 2. 

Vegetation data for stations on the Gamtoos gradient. (A = low altitude station, E = high altitude station) 


90 


Journal of South African Botany 



Station Vegetation type Vegetation structure 1 Dominant species 


Leaf consistence categories in Fynbos Biome and distribution 


91 



92 


Journal of South African Botany 


the summit of the Elandsberg mountains, some 500 m higher. It was poss- 
ible to place transects from approximately sea-level to 500 m in fynbos and 
non-fynbos vegetation. Corresponding stations were located at approxi- 
mately 100 m intervals on each transect and at each station four 100 m 2 sam- 
ples were randomly placed on different aspects (N, S, E, W). All slopes 
were uniformly moderate (8-16°). In each sample a full floristic list was 
made and the projected canopy cover and selected structural features for all 
species were noted. Leaf consistence was determined subjectively. Some 
floristic and structural data for transect stations are shown in Table 2. 

Climate diagrams from three stations along the gradient are shown in 
Figure 1. The extremes of the gradient represent two distinct climates. Han- 
key has the typical valley climate of coastal plain valley bottoms in the south 
eastern Cape. Temperatures are generally higher than adjacent coastal 
mountain and maritime areas. These areas also have greater absolute maxi- 
ma (40 °C) and lower minima (— 1,1 °C) and greater diurnal variation 
throughout the year (see Louw, 1976). Monthly relative humidities are also 
lower, particularly during the warmest part of the day (Louw, 1976). Valley 
areas receive considerably less precipitation than adjacent mountains largely 
because they miss a great deal of post-frontal orographic rain from S and 
SSW winds and frontal rain associated with W and SW winds. Rainfall is 
also highly variable from year to year and may fall in any month although 
long-term averages do indicate spring and autumn maxima (Fig. 1). 




Fig. 1 

Walter- Lieth climate diagrams (mean monthly air temperatures and rainfall) for sta- 
tions on the Gamtoos transect. Temperature data adapted from climatically similar 
nearby station. (Uitenhage (108 m; 52 y) for Hankey, Van Stadens (452 m; 39 y) for 
Otterford). Hankey is representative of Stations A-B, Loerie Lower Forest of Sta- 
tion C, and Otterford of Station E. Data from S.A. Weather Bureau and Depart- 
ment of Forestry. 


The coastal mountain climate (Otterford) is cooler, wetter and less vari- 
able on a daily and annual basis than the valley. Cool marine air keeps the 
temperatures equable although “berg wind” conditions may result in tern- 


Leaf consistence categories in Fynbos Biome and distribution 93 

peratures of 39 C, and frost and snow are occasional in winter (Forestry De- 
partment, pers. comm.). Rainfall is high (900 mm) and falls in every month 
of the year with peaks in spring and autumn (Fig. 1). 

The transects are interpreted as representing gradients of increasing rain- 
fall and decreasing temperature and decreasing climatic variability. The re- 
liability of available soil moisture is likely to improve with increasing alti- 
tude. 

At each station a topsoil sample was collected from a north and south 
slope. Air dried soil samples were analysed for texture, pH and macronu- 
trients by staff of the soil laboratory. Department of Agriculture and Fisher- 
ies, Winter Rainfall Region (Stellenbosch). Soil data are shown in Table 3. 


Results and Discussion 
Consistence determinations 

Results showing leaf characteristics of species are given in Table 4. It is 
important to restrict comparisons among consistence types to leaves of the 
same size category since both degree of sclerophylly and succulence show 
significant positive correlations with leaf size (Camerik and Werger, 1980). 
The degree of sclerophylly ranged from 0,02 to 1,50 gdm -2 (Table 4). Came- 
rik and Wergers’ (1980) values for a tropical high mountain flora ranged 
from 0,29 to 1,19 gdm -2 and they quote Muller Stroll (1947-48) as consider- 
ing species with values of 0,7 gdm -2 and higher as true xerophytes. Within 
our sample all species classed as sclerophyll exceeded this value, except for 
the leptophylls. Within the sclerophyll group, however, leaf size was posi- 
tively correlated with the degree of sclerophylly (r = 0,87 P < 0,01). Ortho- 
phylls, and to a lesser degree fleshy leaves, had lower values while those for 
succulents were comparable to the sclerophylls. 

Degree of succulence was highest for succulents and second highest for 
fleshy leaves. The values of typical succulents are in the range 5,1-14,9 
gdm -2 (Kluge and Ting, 1978) which encompasses the values of both species 
classed as succulents in Table 4. 

Figure 2 indicates that degree of sclerophylly and degree of succulence 
can be used to distinguish leaf consistence categories and lends some cre- 
dence to the subjective scheme. 


Direct gradient analysis 

The relative importance of leaf consistence along the gradient, in fynbos 
and non-fynbos vegetation, is shown in Figure 3. The succulent category in- 
cludes stem succulents (where the stem is the major photosynthesising organ 
and therefore the physiological leaf) and herbaceous succulents. In both fyn 


Table 3. 

Soil data for stations on the Gamtoos gradient. N refers to soil samples from north slope, S at south slope; all samples from top- 
soil (150 mm) 


94 Journal of South African Botany 



photometry. 

3 Total nitrogen determined by the Kjeldahl method. 

4 Available phosphorus on a Bray No. 2 extract. 

5 pH measured in 1 : 2,5 N KC1 (soil to solution ratio). 

6 Colluvial deposits. 


suits are the means of 10 separate leaves ± standard deviations. 


Leaf consistence categories in Fynbos Biome and distribution 


95 


<L> 

OC 


v 

u 


p£3 

o 


cS 

V 



96 


Journal of South African Botany 



Fig. 2. 

Scatter diagram showing relation of subjectively determined leaf consistence for 17 
species to degree of sclerophylly and succulence. S = sclerophyll, O = orthophyll, F 
= fleshy, Sn = succulent, 1 = leptophyll, na = nanophyll, mi = microphyll, me = 

mesophyll. 


bos and non-fynbos the contribution of deciduous species to total cover was 
negligible (0-5 % relative percentage cover). 

Succulents and fleshy leaves 

There is little doubt that the storage of water by plants represents an 
ecological adaptation; the plant is able to overcome periods of water de- 
ficiency with endogenous water reserves (Kluge and Ting, 1978). It would be 
desirable, though not possible in this study, to correlate succulence with the 
crassulacean acid metabolism (CAM) mode of carbon gain since the adapt- 
ive significance of CAM is well understood (Kluge and Ting, 1978). A high 
degree of succulence is, in itself, not a criterion for CAM; only those succu- 
lents which unify the sites of malic acid synthesis, storage and conversion 
within the same cells, can be expected to have CAM (Kluge and Ting, 


Leaf consistence categories in Fynbos Biome and distribution 


97 


FYNBOS NON-FYNBOS 



Fig. 3. 

Importance of leaf consistence categories along an altitudinal gradient in fynbos and 
non-fynbos vegetation in the Gamtoos river valley. Succulents expressed as a per- 
centage of total cover; other categories expressed as a percentage of total woody 
cover. Results are the mean values of four 100 m 2 samples at each station. 


1978). In our study area most of the predominant succulent genera ( Euphor- 
bia , Crassula, Portulacaria, Delosperma, Aloe, Senecio ) are known to pos- 
sess CAM (Mooney, Troughton and Berry, 1977), while another ( Zygophyl - 
lum) is non-CAM (Kluge and Ting, 1978). 

CAM succulents are extremely economical in terms of the ratio of car- 
bon gained to water lost but have associated costs of low rates of carbon 
fixation and are not competitive where soil moisture is not limiting for ex- 
tended periods (Ting and Szareck, 1975; Kluge and Ting, 1978). 

Predictably succulents had maximum cover in the low altitude valley cli- 
mate areas (Fig. 3) with low unpredictable rainfall. Valley regions also ex- 
perience great diurnal temperature fluctuations. It has been shown that opti- 
mum growth of some CAM species occurs under conditions of low night 
(10-15 °C) and high day (25-30 °C) temperatures (Neales, 1973; Osmond, 
Bender and Burris, 1976). 

In fynbos fleshy leaves largely replace succulents at the xeric end of the 
gradient. They are common at corresponding non-fynbos sites with more 


98 


Journal of South African Botany 


fertile soils (Table 3) (cf. Bond, 1981). As yet, the water relations and car- 
bon gaining mechanism of fleshy plants have not been studied. 


Sclerophyll and orthophyll leaves 

There are two major hypotheses to explain the relative fitness of sclero- 
phyll leaves to a particular habitat. The “drought” hypothesis, originally 
stated by Schimper (1903) is best articulated by Orians and Solbrig (1977). 
Sclerophylly has also been interpreted as a response to low levels of nutri- 
ents, particularly nitrogen and phosphorus (e.g. Loveless, 1962). 

Orians and Solbrig (1977) have developed a descriptive cost-income mo- 
del which takes into account water loss and carbon gain as well as morpholo- 
gical and physiological traits that affect these processes. Sclerophyllous 
leaves are interpreted as “high cost-slow profit” structures which are able to 
withstand high negative leaf water potentials through the development of 
energetically expensive supporting structures. These and other biochemical 
changes apparently restrict high photosynthetic rates when water is available 
(see Orians and Solbrig for supporting data) but allow the plant to yield 
profits when more mesophytic leaves have become an energetic liability. 
Sclerophyllous leaves are mostly longlived. 

Orthophyll leaves (“low cost-quick profit”) are cheaper to build and 
maintain per unit surface area and yield profits at faster rates than sclero- 
phyllous leaves, provided soil water potential is low. As negative soil water 
potential builds up they cease to yield profits but maintenance costs con- 
tinue, resulting in a net loss in keeping the leaf (Orians and Solbrig, 1977). 
It is clearly beneficial to drop leaves under these conditions. 

This model predicts that sclerophyllous leaves are at a competitive ad- 
vantage where periods of high soil moisture availability are short and erratic 
(e.g. semi-desert) or do not coincide with the growing season (e.g. mesic 
mediterranean climates (Miller, 1982; Mooney, 1982)). Orthophyll leaves 
will be favoured in tropical savanna, tropical rain forest and temperate 
broadleaf forest climates (Orians and Solbrig, 1977). We would expect, 
therefore, a predominance of sclerophylly at the xeric end of the Gamtoos 
gradients, with orthophylls becoming increasingly important towards the 
higher altitudes. 

With some discrepancies this is the trend in the non-fynbos communities 
(Fig. 3). The trend observed for the fynbos communities does not agree with 
the predictions made by Orians and Solbrig’s (1977) model. Here sclerophyll 
leaves predominate throughout the length of the gradient (Fig. 3). This can- 
not be explained by the soil moisture hypothesis alone; rather we interpret it 
as a response to the low fertility of fynbos soils (Table 3). 

In recent years evidence has accumulated to support the hypothesis that 


Leaf consistence categories in Fynbos Biome and distribution 


99 


a sclerophyllous leaf is the expression of a metabolism found in plants that 
can tolerate low levels of certain nutrients, particularly phosphorus (Love- 
less, 1962; Beadle, 1966; 1968; Small, 1972; 1973; Steurbing and Alberdi, 
1973). Loveless (1962) suggests that the excessive fibre contents typical of 
sclerophyllous leaves is due mainly to an absence of adequate phosphate. 
Thus products of metabolism which otherwise might have formed protein, 
are diverted along alternative metabolic pathways to form other end prod- 
ucts, such as fibre. It is therefore possible to view sclerophylly on infertile 
soils, particularly in areas where soil moisture is not limiting, as a non- 
adaptive feature determined by “biochemical” constraints (Gould and Le- 
wontin, 1979) imposed on the leaf by a metabolism associated with limiting 
amounts of certain nutrients. On the other hand, adaptive advantages can 
be inferred, especially for both arid and infertile environments (Small, 
1973). Small's (1973) hypothesis that there should be a degree of overlap in 
the ecological amplitude of plants adapted to arid and infertile environments 
is rejected for the Gamtoos transects since there are no species common to 
the xeric non-fynbos stations and the fynbos transect. 

In conclusion it appears that patterns in the relative importance of leaf 
consistence in non-fynbos communities on relatively fertile soils, can be in- 
terpreted largely in terms of soil moisture and climatic factors; in fynbos 
communities low levels of soil nutrients may be of overriding importance in 
determining leaf consistence. 


Acknowledgements 

We thank Sue Painting and Wendy Lloyd for valuable laboratory assist- 
ance. R.M.C. collected the Gamtoos field data. This study was funded by 
the CSIR (R.M.C.), the Botanical Research Institute (B.M.C.) and is part 
of the Fynbos Biome Project. W. Bond, M. Werger and W. Westman made 
useful comments on an earlier draft. 

References 

Beadle, N. C. W., 1966. Soil phosphate and its role in molding segments of the Aus- 
tralian flora and vegetation with special reference to xeromorphy and scle- 
rophylly. Ecology 47: 992-1007. 

Beadle, N. C. W., 1968. Some aspects of the ecology and physiology of Australian 
xeromorphic plants. Austr. J. Sci. 30: 348-355. 

Bond, W., 1981. Vegetation gradients in the Southern Cape. M.Sc. thesis. University 
of Cape Town. 

Camerik, A. M. and Werger, M. J. A., 1980. Leaf characteristics of the flora of the 
high plateau of Itatiaia, Brasil. Biotropica 3: 39-48. 

Campbell, B. M., Cowling, R. M., Bond, W. and Kruger, F. J., 1981. Structural 
characterization of vegetation in the Fynbos Biome. South African Nationa 
Scientific Programme Report no. 52. Pretoria: CSIR. 


100 


Journal of South African Botany 


Cowling, R. M. and Campbell, B. M., 1980. Convergence in vegetation structure in 
the mediterranean communities of California, Chile and South Africa. Ve- 
getatio 43: 191-197. 

Gould, S. J. and Lewontin, R. C., 1979. The spandrels of San Marco and the Pan- 
gossian paradigm: a critique of the adaptionist programme. Proc. R. Soc. 
(Lond.) 205: 581-598. 

Kluge, M. and Ting, I. P., 1978. Crassulacean Acid Metabolism. Analysis of an Eco- 
logical Adaptation. Berlin, Heidelberg and New York: Springer. 

Kruger, F. J., 1978. A description of the Fynbos Biome Project. South African Nat- 
ional Programmes Report no. 28. Pretoria: CSIR. 

Louw, W. J., 1976. Meso-climate of the Port JElizabeth-Uitenhage Metropolitan 
area. Technical Paper no. 4. Pretoria: Weather Bureau. 

Loveless, A. R., 1962. Further evidence to support a nutritional interpretation of 
sclerophylly. Ann. Bot. 26: 551-561. 

MacVicar, C. N., de Villiers, J. M., Loxton, R. F., Verster, E., Lamprechts, J. 
J. N., Merryweather, F. R., le Roux, J., van Rooyen, T. H., and von M. 
Harmse, H. J., 1977. Soil Classification. A Binomial System for South Afri- 
ca. Pretoria: Department of Agricultural Technical Services. 

Miller, P. C., 1982. Canopy structure of mediterranean type shrubs in relation to 
the heat and moisture. In: Kruger, F. J., Mitchell, D. T. and Jarvis, J. U. M. 
(eds.), Mediterranean-type ecosystems. The role of nutrients. Berlin: Spring- 
er (in press). 

Mooney, H. A., 1982. Carbon-gaining capacity and allocation patterns of mediterra- 
nean-climate plants. In: Kruger, F. J., Mitchell, D. T. and Jarvis, J. U. M. 
(eds.), Mediterranean-type ecosystems. The role of nutrients. Berlin: Spring- 
er (in press). 

Mooney, H. A. and Harrison, A. T., 1972. The vegetational gradient on the lower 
slopes of the Sierra San Pedro Martir in northwest Baja California. Madro- 
no 21: 439-445. 

Mooney, H. A., Gulmon, S. L. and Parsons, D. J., 1974. Morphological changes 
within the chaparral vegetation type as related to elevational gradients. 
Madrono 22: 281-285. 

Mooney, H. A., Troughton, J. H. and Berry, J. A., 1977. Carbon isotope ratio 
measurements of succulent plants in Southern Africa. Oecologia 30: 
295-305. 

Neales, T. F., 1973. The effect of night temperature on CCb assimilation, transpira- 
tion and water use efficiency in Agave americana L. Aust. J. biol. Sci. 26: 
705-714. 

Orians, G. H. and Solbrig, O. T. , 1977. A cost-income model of leaves and roots 
with special reference to arid and semi-arid areas. Am. Nat. Ill: 677-690. 

Osmond, C. B., Bender, M. M. and Burris, R. H., 1976. Pathways of CO 2 fixation 
in the CAM plant Kalanchoe daigremontiniana. 111. Correlation with 13 C 
values during growth and water stress. Aust. J. Plant Physiol. 3: 787-799. 

Parsons, D. J. and Moldenke, A. R., 1975. Convergence in vegetation structure 
along analogous climatic gradients in the mediterranean climate ecosystems 
of California and Chile. Ecology 56: 950-957. 

Raunkier, C., 1934. The Life Forms of Plants and Statistical Plant Geography. Ox- 
ford: Clarendon Press. 


Leaf consistence categories in Fynbos Biome and distribution 


101 


Schimper, A. F. W., 1903. Plant Geography on a Physiological Basis. Oxford: Cla- 
rendon Press. 

Small, E., 1972. Ecological significance of four critical elements in plants of raised 
Sphagnum peat bogs. Ecology 53: 498-503. 

Small, E., 1973. Xeromorphy in plants as a possible basis for migration between 
arid and nutritionally deficient environments. Bot. Not. 126 : 534-539. 

Steubing, L. and Alberdi, M., 1973. The influence of phosphorus efficiency on scle- 
rophylly. Oecol. Plant. 8: 211-218. 

Ting, I. P. and Szarek, S. R., 1975. Drought adaptation in crassulacean acid metab- 
olism plants. In: N. F. Hadley (ed.), Environmental Physiology of Desert 
Organisms. Stroudsberg: Dowden, Hutchinson and Ross. 

Webb, L. J., Tracey, J. G., Williams, W. T. and Lance, G. N., 1970. Studies in the 
numerical analysis of complex rain forest communities. V. A comparison of 
the properties of floristic and physiognomic structural data. J. Ecol. 58 : 
203-232. 

Werger, M. J. A. and Ellenbroek, G. A., 1978. Leaf size and leaf consistence of a 
riverine forest formation along a climatic gradient. Oecologia 34 : 297-308. 

Whittaker, R. H., 1967. Gradient analysis of vegetation. Biol. Rev. 42 : 207-264. 


MARY GUNN LIBRARY 
NATIONAL BOTANICAL INSTITUTE 
PRIVATE BAG X101 
PRETORIA 0001 

REPUBLIC OF SOUTH AFRICA 



J1 S. Aft. Bot. 49 (2): 103-123 (1983) 


PROTEOID ROOTS IN THE SOUTH AFRICAN PROTEACEAE 


Byron Lamont 

(School of Biology, Western Australian Institute of Technology, Perth, Aus- 
tralia 6102 ) 

Abstract 

The 54 representative species of Proteaceae examined had abundant proteoid 
roots and it is concluded that they are normal components of the root systems of all 
13 genera in southern Africa. Proteoid root terminology is revised and extended in 
view of the predominance of compound types in fynbos vegetation. Their relative 
contribution to the root system may reach up to 80 % total root dry weight but varies 
between individuals and species, and with plant age, season, soil type and location of 
parent roots in the soil profile. While proteoid- root formation is usually dependent 
on soil water availability, the presence is reported of “anomalous” young proteoid 
roots on Leucospermum hypophyllocarpodendron during summer drought. The num- 
ber of rootlets may reach 650 per 10 mm length of parent root providing, together 
with their long root hairs, a substantial increase in surface area available for absorp- 
tion of water and nutrients. There is, however, no simple relationship between abun- 
dance of proteoid roots and plant weight, at least in the first few years of growth. 

UlTTREKSEL 

PROTEOIDE WORTELS IN DIE SUID-AFRIKAANSE PROTEACEAE 

Die 54 verteenwoordigende soorte Proteaceae ondersoek, het volop proteoide 
wortels en word die gevolgtrekking gemaak dat dit normale komponente van al 13 
genera in Suidelike Afrika is. Proteoide terminologie is hersien en uitgebrei in die lig 
van die oorwegend saamgestelde tipes in fynbos-plantegroei. Hulle relatiewe bydrae 
tot die wortelsisteem mag soveel as 80 % droe gewig wees maar dit verskil tussen in- 
dividue en soorte en met die ouderdom van plante, seisoen, en die posisie van die 
ouer-wortel in die grondprofiel. Alhoewel die proteoi'de-wortelvorming afhang van 
die beskikbaarheid van grondwater, word die teenwoordigheid van “afwykende 
jong proteoide wortels by Leucospermum hypophyllocarpodendron gedurende 
somerdroogte vermeld. Die aantal worteltjies kan tot 650 per 10 mm lengte ouer- 
wortel oplewer wat met die lang wortelhare ’n betekenisvolle vermeerdering in 
oppervlakte beskikbaar vir absorbsie van water en voedingstamme is. Daar is egter 
geen eenvoudige verhouding tussen die voorkoms van proteoide wortels en plantge- 
wig nie, ten minste nie gedurende die eerste paar groeijare nie. 

Introduction 

Proteoid roots are dense clusters of rootlets in longitudinal rows (Fig. 1), 
and were first described for the family Proteaceae in Australia by Purnell 
(1960) and elaborated on by Lamont (1972a). Until recently, the South Afri- 


Accepted for publication 11th November, 1982. 


103 


104 


Journal of South African Botany 





Proteoid roots in South African Proteaceae 


105 


can literature on the Proteaceae was devoid of references to these struc- 
tures (e.g. van Staden, 1968) though Vogts (1960) clearly alluded to them in 
her priority-for-research question . why the superficial roots are so ex- 
tremely sensitive and whether the presence of mycorrhizae is necessary to 
continued development.” Proteoid roots are not mycorrhizal but are now 
considered analagous in their function (Lamont, 1982). More recently, Path- 
maranee (1974) and Lamont (1981a), both in Australia, recorded some cul- 
tivated South African species with proteoid roots. In South Africa, Low 
(1980) has also made preliminary records from field excavations in the Cape, 
though proteoid roots were not observed on all plants, and van Staden et al. 
(1981) considered the rootlets of Protea neriifolia which formed in tissue cul- 
ture were proteoid. The aim of this project was to examine representatives 
of the full range of South African Proteaceae for proteoid roots. By quanti- 
tative studies it was also hoped to obtain information on their relative 
morphology, contribution to the root system, occurrence in the field, factors 
controlling their formation and clues to their functional importance to the 
plant. 

Material and Methods 

Root systems were washed free of soil particles, separated into proteoid 
and non-proteoid root portions, and dried at 80 °C for 48 h. For weighing 
purposes, even within compound proteoid roots, parent roots not bearing 
proteoid rootlets were considered non-proteoid. Morphological measure- 
ments were made with vernier calipers, often with the aid of a xlO dissecting 
microscope. Roots for electron microscopy were fixed in 70% alcohol, de- 
hydrated in an alcohol-amyl acetate series, critical-point dried, gold-coated 
and scanned with a Jeol JB 35C electron microscope. 

Nine-month plants referred to in the text were grown in two parts de- 
composed leaf litter: six parts siliceous gravel in plastic pots. They were 
drip-fed daily with tapwater and were 40% shaded in summer. Two-year 
cuttings were grown in two parts compost: one part loam: three parts sand 
in plastic bags. They were watered overhead with tapwater as required and 
kept in the open. Four-month seedlings of Leucadendron laureolum under 


Fig. L 

Proteoid roots as viewed with scanning electron microscope. A, rootlets of mature 
proteoid root showing dense mat of root hairs up to their tips. B, rootlets of young 
proteoid root prior root hair development. C, proteoid rootlets just emerging m 
longitudinal rows from the parent root. D, surface view of mature proteoid re 
showing crowding of rootlets and their close association with soil particles (on e 
left). A-C = Leucadendron salicifolium , D = L. laureolum. Scale: A - 1 mm, - 

= 100 /xm. 


106 


Journal of South African Botany 


the previous treatment were transplanted into 2,5€ plastic pots containing 
either Clovelly sand from Pella via Atlantis or Hutton clay-loam from 
Jonkershoek. They were grown for a further 22 wk in a plastic greenhouse, 
watered with tapwater, and three random replicates of each were selected 
for root analysis. 

Results and Discussion 
Taxonomic distribution 

Table 1 shows that all specimens of all (54) species examined possessed 
proteoid roots (nomenclature follows the latest revision available). This in- 
cluded two Aulax species (three in genus), one Brabejum (one), three Dia- 
stella (five), two Faurea (five), ten Leucadendron (81), six Leucospermum 
(47), two Mimetes (16), one Orothamnus (one), two Paranomus (17), 12 
Protea (83), eight Serruria (50), two Sorocephalus (11) and three Spatalla 
(20). 

In association with what is already known about the proteoid-root bear- 
ing propensity of the sub-families Proteoideae (12 genera in South Africa) 
and Grevilleoideae (one genus in South Africa) (Lee, 1978; Lamont, 
1981a), the results here suggest that proteoid roots are normal components 
of the root systems of all taxa in (southern) Africa. This is in contrast with 
Australia, where there are five genera lacking proteoid roots, especially in 
the primitive Persoonioideae. (The claim that Agastachys (Proteoideae) 
lacks proteoid roots (Lee, 1978) needs further investigation). 

Expanded terminology 

Following Purnell (1960), simple proteoid roots comprise a single cluster 
of rootlets (Fig. 2a). They are sometimes produced sequentially (in “series”) 
along a parent root during the one growing season (Lamont, 1972a). Com- 
pound proteoid roots are defined here as a group of three or more clusters 
of rootlets borne on three or more parent roots (in “parallel”). The com- 
pound root therefore refers to a cluster of simple proteoid roots (Purnell, 
1960: Fig. 2b). Clusters of rootlets are considered here to belong to separate 
compound proteoid roots if they are separated by more than one lateral root 
not bearing clusters of rootlets. Second order parent roots may arise from 
the axis of a primary proteoid root borne along the first order (original) par- 
ent root. This is defined here as a “closed” compound proteoid root. If the 
secondary proteoid roots arise from a non-proteoid portion of the parent 
root, the compound proteoid root is regarded as “open” (Fig. 2b). If only 
some of the second-order parent roots arise from a proteoid axis it is “in- 
completely-closed”. If a primary proteoid root is present (usually terminal) 
but none of the second-order parent roots arise from it, the compound pro- 


Proteoid foots in South African Protcaccac 


107 




a. Part of a new season’s lateral branch of Serruria florida with three simple proteoid 
roots, b. The same for Leucospermum formosum showing an “open” compound root 
with 28 second order (2°) parent roots and 111 simple proteoid root components 
(middle portion omitted). Note that all but two of the third order (3°) parent roots 
are “closed”. Not drawn to scale. 


108 


Journal of South African Botany 


teoid root is “incompletely-open”. Tertiary and (the extremely rare) higher 
order proteoid roots are usually closed with respect to their parent roots, 
but otherwise the same terminology applies. For example, the compound 
proteoid roots of Leucospermum formosum are open or incompletely-open, 
with closed or incompletely-closed tertiary proteoid roots, and closed 
quaternary and quinternary proteoid roots (Fig. 2b). 

Morphometries 

Returning to Table 1, all plants possessed at least some simple proteoid 
roots, with those of the genera containing the smallest plants, Serruria, 
Sorocephalus and Spatalla, remaining essentially simple throughout the life 
of the plant. Compound proteoid roots, usually open, were common in the 
other genera and contributed increasingly to the root system as plants in- 
creased in age (e.g. Brabejum stellatifolium, Table 1). Only Leucospermum 
and Mimetes were well represented by partly-closed, as well as open, com- 
pound proteoid-roots in mature plants, while most proteoid roots of Oro- 
thamnus were closed compound, though they were sparsely branched. The 
maximum number of second-order parent roots per compound proteoid root 
was 40 ( Leucospermum conocarpodendron) , with Protea repens having up to 
35. Branching of compound proteoid roots was least in Spatalla, Soroce- 
phalus, Serruria and Faurea. Individual proteoid roots tended to be smallest 
in Spatalla, Paranomus and Faurea (<35 mm long, <15 mm wide) and larg- 
est in Leucadendron and Leucospermum {<15 mm long, <35 mm wide). 
Older plants tended to have larger individual proteoid roots (e.g. B. stellati- 
folium, Leucadendron laureolum) . The longest, simple proteoid root noted 
was 75 mm ( Leucospermum conocarpodendron) and the shortest was 
1,5 mm {Spatalla racemosa and O. zeyheri). The widest was 34 mm {Leuco- 
spermum parile) and the narrowest 1,4 mm {Spatalla incurva ). 


Table 1. 

Occurrence and morphology of proteoid roots in 54 species of Proteaceae indigenous 
to southern Africa. 




Description of proteoid roots 

Species 

Specimens 

Type + 

No. of 
2° rts* 

Length 

(mm) 

Width 

(mm) 

Aulax cancel lata 

9 mth seedlings, sand-hu- 
mus mix 

s,o/c 

— 

5-30 

— 

A. umbellata 

9 mth seedlings, sand-hu- 
mus mix 

50 

— 

5-40 

— 


Mature shrub, sandstone, 
Kogelberg 

s,o,c 

10 

3-44 

3-14 


Proteoid roots in South African Proteaceae 


109 




Description of proteoid roots 

Species 

Specimens 

Type + 

No. of 
2° rts* 

Length 

(mm) 

Width 

(mm) 

grabejum stellati- 
folium 

2 yr seedlings, shale, 
Kirstenbosch 

S,0 

9 

5-27 

7-15 


7 m tree, shale, Kirsten- 
bosch 

o 

20 

6-43 

6-20 

Diastella buekii 

2 yr cutting, sandy loam, 
farstenbosch 

s,o, 

o/c,c 

10 

3,5-35 

6,5-13 

D. proteoides 

Mature shrubs, fine sand, 
Pella 

s,o,o/c 

8 

5-39 

5-22 

D. serpyllifolia 

3,5 yr shrublet, rocky 
sandstone, Bot River 

s,c,o/c 

7 

6,5-29 

4-13,5 

Faurea mac- 
naughtonii 

2 yr cutting, sand-humus 
mix 

s 

0 

10-16 

5-12,5 


1,4 m seedling, humic 
loam, Knysna forest 

s,o/c 

2 

3-19 

4-15 

F. speciosa 

5 yr shrub, sandy loam, 
Kirstenbosch 

s,o 

7 

2-21 

3,5-12 

Leucadendron 

argenteum 

3 m tree, sandy loam, 
Kirstenbosch 

s,o 

28 

3-66 

5-13 

L. discolor 

9 mth seedlings, sand-hu- 
mus mix 

s 

0 

5-20 


L. eucalyptifolium 

9 mth seedlings, sand-hu- 
mus mix 

5,0 

— 

5-30 



6 yr shrub, laterite 0 

s,o 

— 

— 

— 

L. floridum 

2 yr cuttings, sand-humus 
mix 

5,0 

6 

2,5-27 

2,5-17 

L. galpinii 

3 yr shrub, coarse sand 

5,O 0 

— 

— 

— 

L. laureolum 

10 mth potted seedlings, 
fine sand, Pella 

5,0, 

c,o/c 

12 

3,5-40 

2,3-18 


13 yr shrub, rocky sand- 
stone, Bot River 

s,o,o/c 

24 

3-60 

4-20 

L. muirii 

3 yr shrub, coarse sand 0 

0 

— 

— 

— 

L. salicifolium 

10 mth seedlings, sand- 
humus mix 

s 

0 



L. salignum 

2 yr seedlings, sandstone, 
Bainskloof 

s 

0 

30 

10 


5 yr regrowth, fine sand, 
Pella 

s,o 




Leucospermum 

cordifolium 

2 yr seedlings, coarse 
sand 0 

s,o/c 




2 yr cuttings, sand-humus 
mix 

o,c, 

o/c 

7 

6-31 

4-13 

L. conocarpoden- 
dron 

2,0 m shrubs, stony sand, 
base Table Mt 

S ,0, 

c,o/c 

40 

2,5-75 

3,5-15 

L. formosum 

2 yr cuttings, sand-humus 
mix 

o,c, 

o/c 

27 

2,5-21 

3-11 

L. hypophyllo- 
carpodendron 

Mature rhizomes, fine 
sand, Pella 

o,c, 

o/c 

10 

5-35 

5-29 


110 


Journal of South African Botany 




Description of proteoid roots 

Species 

Specimens 

Type + 

No. of 
2° rts* 

Length 

(mm) 

Width 

(mm) 

L. parile 

Mature shrubs, fine sand, 
Pella 

s,o,o/c 

20 

29-54 

10-34 

L. truncatulum 

2 yr seedlings, coarse 
sand 0 

s,o,c 

— 

10-30 

6 

Mimetes cucullatus 

2 yr cuttings, sand-humus 
mix 

s,o, 

c,o/c 

8 

4-22 

3,5-15 

M. hirtus 

2 yr cuttings, sand-humus 
mix 

s,o,c 

9 

4-31 

4-17 

Orothamnus zeyheri 

2, 5 yr seedling, potted 
sand 

s,o,c 

5 

1,5-31 

6-13 

Paranomus reflexus 

600 mm cutting, loamy 
sand, Kirstenbosch 

s,o,c 

7 

4-22 

2-8 

P. spicatus 

2 yr cuttings, sand-humus 
mix 

s,o,o/c 

7 

2-22 

4-9 

Protea acaulos 

Mature rhizomes, fine 
sand, Pella 

s,o,o/c 

7 

8-34 

7-23 

P. nitida 

2 m tree, sandstone, 
Bainskloof 

o 

4 

10-25 

10-15 

P. aurea 

3 yr shrub, coarse sand 0 

o 

_ 

_ 

_ 

P. burchellii 

2 m shrub, fine sand, Pel- 
la 

2 yr seedlings, sand-hu- 
mus mix 

c 

6 

50 

30 

P. cynaroides 

s,o,c 

9 

2-6 

4-13 

P. discolor 

3 yr seedlings, coarse 
sand 0 

2 yr seedlings, sand- 
stone, Bainskloof 

s,o 

— 

— 

— 

P. laurifolia 

s 

0 

— 

— 

P. magnifica 

3 yr seedlings, coarse 
sand 

o 

— 

— 

— 

P. neriifolia 

2,5 m shrub, fine sand, 
Kirstenbosch 

s,o,c 

17 

2-36 

3-17 

P. repens 

1 m shrub, fine sand, Pel- 
la 

18 mth seedlings, coarse 
sand 0 

12 mth seedling, coarse 
sand 0 

s,o,o/c 

35 

5-53 

2-17 

P. speciosa 

o 

— 

— 

— 

P. stokoei 

5,0 

— 

— 

— 

Serruria aitonii 

2 yr cuttings, sand-humus 
mix 

s,o, 

c,o/c 

13 

4-30 

3-19 

S. burmannii 

Flowering shrublet, fine 
sand, Pella 

s 

0 

25 

8 

S. ciliata 

2 yr cuttings, sand-humus 
mix 

5,0 

7 

3-19 

3-11 

S. elongata 

3 yr cuttings, sand-humus 
mix 

s 

0 

3-15 

2-10 

S. florida 

2 yr cuttings, sand-humus 
mix 

5,C,0/C 

6 

7-32 

4-13 


Proteoid roots in South African Proteaceae 


111 




Description of proteoid roots 

Species 

Specimens 

Type + 

No. of 
2° rts* 

Length 

(mm) 

Width 

(mm) 

5. pedunculata 

3 yr shrub, coarse sand 0 

S 

0 





S. simplicifolia 

2 yr seedlings, fine sand, 
Pella 

S 

0 

10-20 

5 

S. vallaris 

2 yr cuttings, sand-humus 
mix 

S 

0 

3-29 

6-16 

Sorocephalus lana- 
tus 

2 yr cutting, brown sand, 
Kirstenbosch 

S 

0 

4-17 

2-5 

S. capitatus 

1,5 yr cuttings, sand-hu- 
mus mix 

5,0 

4 

3,5-36 

3,5-17 

Spatalla incurva 

2 yr cutting, brown sand, 
Kirstenbosch 

5,0 

— 

2-23 

1,4-7 

S. parilis 

2 yr cutting, brown sand, 
Kirstenbosch 

5,0 

— 

2-19 

1,5-6 

S. racemosa 

5 yr shrublet, rocky sand- 
stone, Bot River 

5,0 

3 

1,5-20 

1,5-9 


+ S = simple proteoid roots, O = open compound, C = closed compound, O/C = 
incompletely open or closed, letter in italics = most proteoid roots of this type 
0 Cultivated plants in Perth, Western Australia 

* Maximum number of second-order roots within a compound proteoid root bear- 
ing simple proteoid roots 

Note: Length and width measurements refer to simple proteoid roots, even when 
these are components of compound proteoid roots. 


Ontogenetically, and perhaps phylogenetically (Johnson and Briggs, 
1975), the results suggest that compound proteoid roots are more ad- 
vanced” than simple types. Compound proteoid roots with many branches 
are more advanced than those with few branches. Closed are more advanced 
than incompletely-closed, which are more advanced than incompletely- 
open, which are more advanced than open. Similarly, compound proteoid 
roots with higher order clusters, are more advanced than those with lower 
order clusters. 

All proteoid roots were made up of two (some Paranomus reflexus roots) 
to six (some Leucadendron argenteum roots) longitudinal rows of rootlets, 
as found for the genus Hakea in Australia (Lamont, 1972a). Proteoid root- 
lets varied in average width from 80 /am ( Sorocephalus lanatus) to 250 /am 
(e.g. Faurea speciosa). These dimensions are much less than those for root 
lets of Protea neriifolia developed in tissue culture (1,5-5 mm) and claimed 
to be proteoid (van Staden et al., 1981). Their approximate length can be 


112 


Journal of South African Botany 


determined by dividing width of proteoid roots in Table 1 by two. Most 
rootlets were in single rows (Fig. 1C), but those of Mimetes cucullatus and 
M. hirtus were in double rows due to compaction displacing alternate root- 
lets sideways (illustrated in Lamont, 1972a and to some extent in Fig. IB). 
Leucospermum hypophyllocarpodendron was unusual in possessing some 
proteoid roots with non-contiguous “incomplete” rootlets (Table 5). The 
number of rootlets per unit axis length varied from 60 (L. argenteum) to 140 
(5. lanatus) per 10 mm row, depending on diameter and arrangement of the 
rootlets (Table 5). Including all rows, this gives about 200 rootlets/10 mm 
axis length for ten-month seedlings of Leucadendron laureolum to 650 
rootlets/10 mm axis length for typical proteoid roots of rooted cuttings of 
M. cucullatus. The root hairs on proteoid rootlets were often much longer 
than those on non-proteoid roots (compare Fig. 1C and D), ranging from 
250 /xm ( Serruria aitonii) to 725 /xm (L. hypophyllocarpodendron) . 

In an experiment with Leucadendron laureolum (Lamont, unpubl.), ma- 
ture proteoid roots had 2,6 times the surface area per unit dry weight as ad- 
jacent non-proteoid roots, omitting any contribution by root hairs. The root 
hairs on proteoid rootlets were three times the length of those on non-pro- 
teoid roots (though of similar width and density). This gave an additional 
5,84 mm 2 of surface per 1 mm 2 of epidermal surface, 2,6 times that of equiva- 
lent non-proteoid roots. The total root surface of mature proteoid roots per 
unit weight was consequently 5,4 times that of mature non-proteoid roots. 
This is consistent with Jeffrey’s (1967) claim of 4,5 times for Banksia ornata, 
though the dense cover of root hairs on the proteoid roots of other species 
(e.g. Brabejum stellatifolium, Faurea macnaughtonii, Protea cynaroides ) 
would greatly increase this figure. 


Contribution to the root system 

Nine months after germination, pot-grown B. stellatifolium had no pro- 
teoid roots, while they were abundant in Aulax cancellata, A. umbellata, 
Leucadendron eucalyptifolium and L. laureolum (Table 2). The delay in 
proteoid root initiation in B. stellatifolium was no doubt due to the large 
store of nutrients in the still-intact drupe. The smaller cotyledons on the 
other species, except Aulax umbellata , had already decayed. That com- 
mencement of proteoid root initiation is not necessarily tied to cotyledon 
abscission however (Lamont, 1972a) is confirmed by A. umbellata (and Pro- 
tea cynaroides. Table 4), which had the greatest concentration of proteoid 
roots. As indicated by Leucadendron, there was as much variation in con- 
centration of proteoid roots in the root systems between species in a given 
genus as there was between genera (see also Table 4). At this early stage in 
growth, there was also much variation in proteoid root production between 


Table 2. 

Production of proteoid roots by six pot-grown species of Proteaceae in relation to shoot and total root growth, nine months 
after germination. Results are mean ± standard deviation for six plants, except Leucadendron discolor (twelve plants). Signi- 
ficant curvilinear correlations between number of proteoid roots and plant weight are indicated. 


Proteoid roots in South African Proteaceae 


113 



$ 



114 


Journal of South African Botany 


5 +i 


OC4 



cor-i^fcococococo 

l-t N 

O O O O O O O O 

D-i) 

+i +i +i +i +i +i +i +i 

£ 2 c 

^ CX 

r-H SO CO *"1 CO i-H 00 t> 

H o' T-H ^ T— T T-H (N 


rHsocNsooor^inco 

% 

roteoid 
rt wt/ 
:otal rt 
wt 

vo r<i oi cl \6 1-4 
+1 +1 +1 +1 +1 +1 +1 +1 
Nh'CrO'ChOO 

Q. 

OfNri'O^tOM^o 
n t-n m h n ^ -h 

33 


No. 

proteo 

rts/g 

total 

*wt 

176 ± 
249 ± 
18 ± 
122 ± 
121 ± 
272 ± 
92 ± 
39 ± 

Ratio 

A/B 

00 <0 (N CO 00 (N os 
in so" co (N co co" tj- 


•^t’^t'^fOSOOSO'^'CN 

1-^fOfOOOinn 

£ CQ 

o o o o o o o o 

o ^ 

+1 +1 +1 +1 +1 +1 +1 +1 

Oc<‘>'0<Nmor--rr> 
r--_ Tt o o o oo ^ 
0 > 1 —* t~h i— ( t-h o *~h C> 

% < 

t^'CO^HCOOO'^-^H 

OO t-H SO i-H t-H 00 SO 

o' i-H © H o' o" © o 

o“ 

+1 +1 +1 +1 +1 +1 +1 +1 

She 

(g) 

TiOOhOOhOOH 

<o o in as (N (N © 

co r- so co" cn" r- <n" 


Leucadendron floridum 
Leucospermum formosum 
L. oleifolium 
Mimetes hirtus 
Paranomus spicatus 
Protea cynaroides + 
Serruria aitonii 
S. vallaris 


Ten month seedlings with cotyledons still present; other species grown from cuttings. 


Proteoid roots in South African Proteaceae 


115 


replicates, as suggested by the high standard deviations. Proteoid root con- 
centration in the rooted cuttings from mature plants (Table 4) was much 
more stable. The greater the number of proteoid roots in the root system of 
replicates, the higher the proportion of compound ones, suggesting physiol- 
ogical advancement of these plants (see earlier comments on Table 1). 

Correlations (linear, logarithmic, exponential, power) between dry shoot 
weight and number of proteoid roots per seedling, were only significant for 
two of the six species. In functional terms, then, increased occurrence of 
proteoid roots cannot be held responsible for instances of greater plant 
growth in four of the six species. The two positive correlations can even be 
interpreted as the reverse — larger plants, under otherwise identical con- 
ditions, are more likely to have proteoid roots (Lamont, 1981a). 

Advantage was taken of the absence of proteoid roots in six plants of 
Leucadendron discolor to compare their growth against six plants with them 
(Table 3). Again, those with proteoid roots were not larger than those with- 
out them. This is no doubt due, at this young stage, to lack of dependence 
on external supply of nutrients and too recent formation of proteoid roots. 
Variations in seed size, date of germination and soil attributes (nutrients, air 
and water availability, stimulatory bacteria, see Lamont, 1981a) between 
replicates could explain the apparent random fluctuations in occurrence of 
proteoid roots between plants. 

Of seven species of rooted cuttings, only Serruria aitonii, S. vallaris and 
Leucospermum oleifolium had <100 new season’s proteoid roots/g total root 
weight (Table 4). The root systems of cuttings can be regarded as equivalent 
to new season’s growth of mature plants, since these also are invariably of 
adventitious origin (Lamont, 1972a). L. oleifolium was the only species of 
the seven in flower (mid-March), had the fewest proteoid roots and yet the 
highest shoot/root ratio. Reproductive growth may well compete with root 
growth in this species and phenological studies might be informative. 
L. formosum on the other hand had greatest concentration of proteoid 
roots, up to 440/g total dry root weight in one replicate. This compares 
favourably with the other highest records of proteoid root abundance. 354 
for one P. cynaroides seedling (Table 4), 557 for portion of the root system 
of an 18-month Leucadendron laureolum in humus-rich soil (Lamont, 
unpubl.) and 578 for a surface lateral of a 3 m high L. argenteum at Kirsten- 
bosch. 

Proteoid roots were not weighed individually, but ranged from an aver 
age weight of 0,5 mg ( Leucospermum formosum, mostly immature) to 
3,0 mg ( Serruria vallaris) throughout the root system (Table 4). For com 
parison, proteoid roots in Leucadendron laureolum seedlings of simi ar age 
weighed on average 7,3 mg in Clovelly sand (Table 7). There was some ten 
dency for species with a greater number of proteoid roots to have sma er 


116 


Journal of South African Botany 


ones on average (Table 4). This was apparently due to an increased propor- 
tion of compound proteoid roots with higher order (i.e. smaller) laterals 
bearing most of the (consequently smaller) simple proteoid root components 
(see Fig. 2b and Table 1). 

The contribution of living proteoid roots by weight to the total root sys- 
tem varied from 1 % (L. oleifolium) to 37 % ( Mimetes hirtus) (Table 4). 
Proteoid roots (including dead) accounted for 80 % of the dry weight of the 
surface lateral of the specimen of L. argenteum noted above. If proteoid 
roots increase the efficiency of nutrient uptake (Jeffrey, 1967; Malacjzuk 
and Bowen, 1974; Lamont, 1982) then a positive correlation between % 
weight of proteoid roots in the root system and shoot/root ratio might be ex- 
pected. That is, as the % contribution of proteoid roots increases, propor- 
tionately more shoot weight should be supported. No such simple, mechan- 
ical relationship is evident between replicates or species in Table 4. The 
reason is that growth is as much dependent on initial plant nutrient status 
and species recycling ability (Grundon, 1972) as on root morphology. 


Factors affecting formation 

Age. After nine months, all pot-grown seedlings of Aulax umbellata, 
Leucadendron eucalyptifolium and L. laureolum had proteoid roots in abun- 
dance (Table 2), but only half of L. discolor and no B. stellatifolium had 
them. Study of the two Leucadendron species over time (Fig. 3) shows ab- 
sence of proteoid roots up to 17 weeks. By about 25 weeks all seedlings pos- 
sessed them. L. laureolum shows a power function increase in number of 
proteoid roots with increase in shoot growth (r sig. at 1 % level). As dis- 
cussed earlier, it is not possible to unravel cause and effect from such data. 
It is clear, however, that the older the plant the more likely it is to have pro- 
teoid roots, and that by the second year of growth they are standard compo- 
nents of the root systems of these species (also see Table 1). 

Soil type. Nine-month seedlings of Leucadendron laureolum grown in 
Clovelly-type soil (fine sand) had almost 50% more, and larger, proteoid 
roots than those in Hutton-type soil (clay-loam) (Fig. 4). This was partly at- 
tributable to the fact that root weight in Clovelly soil was twice that in the 
Hutton, despite the fact that the Clovelly soil had only one-third the level of 
extractable phosphate as the Hutton (C. Brown, unpubl.). It is possible that 
phosphate was at inhibitory levels in the Hutton, although in other species 
proteoid root production ceases altogether at levels which inhibit “normal” 
root growth (Lamont, 1972b). In view of the high water content of the 
potted Hutton soil (Lamont, unpubl.), the results are more consistent with 
the effects of mild waterlogging on root growth as demonstrated for other 
proteaceous species (Lamont, 1976). 


Table 5. 

Distribution of recently-emerged proteoid rootlets along the parent axis. Double rows result when alternate rootlets are dis- 
placed laterally as they emerge. Contiguous rootlets lack gaps between them within the longitudinal row. Incomplete rows have 

gaps exceeding diameter of rootlets within the axis. 


Proteoid roots in South African Proteaceae 


117 



118 Journal of South African Botany 



Fig. 3. 

Presence of proteoid roots in relation to age of seedlings of Leucadendron laureolum 
(a, ten plants at each harvest) and L. eucalyptifolium (b, seven plants). Mean ± stan- 
dard errors are given. O-shoot dry wt, ■-% plants with proteoid roots, D-no. pro- 
teoid roots per plant. 


Soil profile distribution. Examination of many plants in a range of 
soil types (Table 1) repeatedly confirmed that the proteoid roots were con- 
centrated in the uppermost 100 mm of soil, especially in the decomposing 
litter and humus layers beneath the canopy of the plant, as found in Austra- 
lia (Lamont, 1973). The abundance of compound proteoid roots in larger 
and older plants (Table 1) resulted in a distinct mat of proteoid roots just 
under the raw litter (e.g. Leucadendron salignum and Protea burchellii at 
Pella) though this only remained continuous under the plant canopy, unlike 
Banksia in Australia which may extend between plants (Jeffrey, 1967). The 
results for adult plants of Leucospermum parile at Pella will be described 
elsewhere (S. Jongens-Roberts and D. T. Mitchell, unpubl.), but prelimi- 
nary data for two-year-old plants (Table 6) show that the pattern is estab- 
lished early. In this case, there was no decomposing litter layer and 90 % of 
all proteoid roots were concentrated in the uppermost 50 mm of soil. Both 
proteoid root number and weight fell away exponentially with depth, with 


Proteoid roots in South African Proteaceac 



Fig. 4 

Dimensions of proteoid roots of L. laureolum in one plant grown in Clovelly sand 
(O) and one in Hutton clay-loam (•) for nine months. Note the essentially linear re- 
lationship between axes and rootlet lengths, and the paucity of large proteoid roots 
in the Hutton soil. Part of the wide variation about the best-fit line is due to inclusion 
of immature proteoid roots. 


120 


Journal of South African Botany 


none below 150 mm. This pattern was in no way related to the distribution 
of lateral roots which remained fairly constant down to 600 mm, nor to soil 
water availability which increased below 300 mm (Table 6). There was a lin- 
ear, rather than exponential, drop in soil organic matter with depth but, to- 
gether with the nutrients released from the fire that initiated germination, 
the proliferation of proteoid roots at the surface can be related to enhanced 
nutrient availability here (Lamont, 1973). 

Seasonality. The root systems of thirteen species examined in the 
field (Kirstenbosch, Bainskloof, Kogelberg Mts and Pella, see Table 1) dur- 
ing March 1980 usually only possessed dead proteoid roots. The rootlets 
were brown, broken off or lacked cortices or root hairs. The exceptions 
were plants of Leucospermum hypophyllocarpodendron examined in Clovel- 
ly sand at Pella on 3rd and 18th March, 1980. Some of the adventitious roots 
arising from the rhizomes possessed young and mature compound proteoid 
roots. The rootlets were white and densely-covered with root hairs (Tables 
1, 5). The rhizosphere soil had a water content of 3,34% (above field ca- 
pacity of 2,4%), while the surrounding soil had a water content of 0,34% 
(below permanent wilting point of 1,8%). One explanation is that dew 
rolled down the vertically-oriented leaves, over the rhizome and down the 
parent root to create a locally-moist environment for root growth. In view of 
the small surface area of these leaves, the hydrophobicity of this soil when 
dry, and the unreliability of such a source for ongoing root growth, this 
possibility seems remote. Active roots are quite capable of releasing water 
on the other hand (e.g. Lamont, 1981b), and a more likely explanation is 
that water was provided from the sub-soil via the massive taproot and re- 
leased into the rhizosphere. By maintaining a moist rhizosphere these 
“anomalous” summer proteoid roots may serve to prolong nutrient release 
from the organic particles and take advantage of “out-of-season” showers 
(Lamont, 1982). 

Following the first substantial autumn rains at Pella both “normal” and 
proteoid root growth commenced in Leucospermum parile at least (Table 7). 
By mid-June there was vigorous proteoid root growth at the same site by 
Diastella proteoides (flowering), Leucadendron salignum, Protea repens and 
P. burchellii and to a lesser extent Serruria simplicifolia. October appeared 
to be the critical month for cessation of proteoid root growth (Table 7). By 
mid-October, the surface soil around fifth-year plants of Protea repens and 
P. burchellii was dry and all roots were moribund. Under 20-year stands of 
these species however, with a thick layer of litter and canopies to the 
ground, the soil was still moist and young and mature proteoid roots were 
abundant. By late October proteoid root growth here had also ceased. How- 
ever, excavations around P. neriifolia, Leucadendron argenteum and Leu- 
cospermum conocarpodendron at Kirstenbosch over summer showed growth 


Table 6. 

Distribution of roots through the soil (Clovelly-type) profile of two-year-old plants of Leucospermum parile at Pella (via Atlan- 
tis) on 2 June 1980 


Proteoid roots in South African Proteaceae 


121 




Organic 
matter (%) 

On vO rf oo 

H t-T r-T O' O' 

Soil water 
content (%) 

Q G 

00 — ^ CN <N 00 

o on so vo -t 
ri i-H i-T t-T rJ cn 

Wt of 
tap root 

F- O — ' F-- ^ 

oo cn ro cn rn © 
o' o' o' o' o' o' 

Wt of 
laterals 

F F 00 O (N r, 

Wt of 

proteoid roots 

0,31 

0,03 

0,02 

0,00 

0,00 

0,00 

No. of 

proteoid roots 

O 00 o o o 

OS 

Soil* 

Depth (mm) 

0-50 

50-100 

100-150 

150-300 

300-450 

450-600 


o o 


E C8 





no growth, + = moderate growth, ++ = maximum growth 
a few moist clusters present. 


122 


Journal of South African Botany 


of both types of roots continued provided the soil remained moist (e.g. 
5,9% water content recorded beneath L. conocarpodendron). Examination 
of Leucadendron laureolum, Diastella serpyllifolia and Spatalla racemosa at 
Bot River on December 10 showed all surface roots to be dormant. Clearly, 
where surface water availability is highly seasonal so also is root growth, in- 
cluding production of proteoid roots, as previously demonstrated in a re- 
lated mediterranean ecosystem (Lamont, 1976). 

Acknowledgements 

Bill Storm van’s Gravesande propagated many of the plants especially 
for this project (Tables 1-3, Fig. 3). Kirstenbosch Botanic Gardens pro- 
vided additional plants (Tables 1, 4) and John Winter and John Rourke are 
thanked for their support. Gary Brown helped me analyse the root systems 
of his L. laureolum plants, Gert Brits (Tygerhoek Experimental Farm) al- 
lowed me to inspect the roots of his Orothamnus plants and William Bond 
and Hans van Daalen (Saasveld Forestry Research Station) took me to Fau- 
rea macnoughtonii . Ruth, Walter and Robert Middelmann provided gener- 
ous hospitality at Bot River. Eugene Moll, Margie Jarmen, Derek Mitchell 
and all other members of the Fynbos Biome Project at the Botany Depart- 
ment, University of Cape Town are thanked for their support. Liz van der 
Pennen provided technical expertise on the SEM at WAIT. Research funds 
were provided by the CSIR and University of Cape Town (1980, with E. 
Moll). The project was undertaken during study leave granted by WAIT. 

References 

Grundon, N. J., 1972. Mineral nutrition of some Queensland heath plants. J. Ecol. 
60: 171-181. 

Jeffrey, D. W., 1967. Phosphate nutrition of Australian heath plants. 1. The import- 
ance of proteoid roots in Banksia (Proteaceae). Aust. J. Bot. 15 : 403-411. 
Johnson, L. A. S. and Briggs, B. G., 1975. On the Proteaceae — the evolution and 
classification of a southern family. J. Linn. Soc. Bot. 70 : 83-183. 

Lamont, B., 1972a. The morphology and anatomy of proteoid roots in the genus Ha- 
kea. Aust. J. Bot. 20 : 155-174. 

Lamont, B., 1972b. The effect of soil nutrients on the production of proteoid roots 
by Hakea species. Aust. J. Bot. 20 : 27^-0. 

Lamont, B., 1973. Factors affecting the distribution of proteoid roots within the root 
systems of two Hakea species. Aust. J. Bot. 21 : 165-187. 

Lamont, B., 1976. The effects of seasonality and waterlogging on the root systems of 
a number of Hakea species. Aust. J. Bot. 24 : 691-702. 

Lamont, B. B., 1981a. Specialized roots of non-symbiotic origin. In: Specht, R. L. 

(ed.), Heathlands and Related Shrub lands of the World. B. Analytical Stu- 
dies. Amsterdam: Elsevier, pp. 183-195. 

Lamont, B., 1981b. Morphometries of the aerial roots of Kingia australis (Liliales). 
Aust. J. Bot. 29 : 81-96. 


Proteoid roots in South African Proteaceae 


123 


Lamont, B., 1982. Strategies for maximising nutrient uptake in two mediterranean 
ecosystems of low nutrient status. In: Kruger, F. J., Mitchell, D. T. and 
Jarvis, J. M. (eds.), Mediterranean-type Ecosystems: The Role of Nutrients. 
Berlin: Springer-Verlag (in press). 

Lee, H. M., 1978. Studies of the family Proteaceae. II. Further observations on the 
root morphology of some Australian genera. Proc. R. Soc. Viet. 90' 
251-256. 

Low, A. B., 1980. Preliminary observations on specialized root morphologies in 
plants of the Western Cape. S. Afr. J. Sci. 76 : 513-516. 

Malajczuk, N. and Bowen, G. D., 1974. Proteoid roots are microbially induced. 
Nature, Lond. 251 : 316-317. 

Pathmaranee, N., 1974. Observations on Proteoid Roots. M.Sc. thesis. University of 
Sydney, N.S.W. 123 pp. 

Van Staden, J., 1968. The effect of trace element deficiencies on the Bot River Pro- 
tea ( Protea compacta R.Br.) in water culture. // S. Afr. Bot. 34 : 397-400. 

Van Staden, J., Choveaux, N. A., Gilliland, M. G., McDonald, D. J. and Da- 
vey, J. E., 1981. Tissue culture of South African Proteaceae. 1. Callus and 
proteoid rootlets formation on cotyledonary explants of Protea neriifolia. S. 
Afr. J. Sci. 77: 493-495. 

Vogts, M., 1960. The South African Proteaceae: the need for more research. S. Afr. 
J. Sci. 56 : 297-305. 




J1 S. Afr. Bot. 49 (2): 125-142 (1983) 


POLLINATION OF MIMETES HIRTUS (PROTEACEAE) BY CAPE 
SUGARBIRDS AND ORANGE-BREASTED SUNBIRDS 

Brian G. Collins* 

(Percy FitzPatrick Institute of African Ornithology, University of Cape 
Town, Private Bag, Rondebosch 7700, R.S.A.) 


Abstract 

Promerops cafer (Cape Sugarbird) is a major pollen vector for Mimetes hirtus, 
brushing its head and throat against pollen presenters and stigmata as it forages for 
nectar. P. cafer is attracted to partially-opened inflorescences, in particular, by virtue 
of the large nectar standing crops present. Maximum pollen availability and viability 
coincide with most visits by P. cafer, and result in a relatively high degree of seed 
set. Nectarinia violacea (Orange-breasted Sunbird) visits M. hirtus occasionally, but 
does not transfer significant pollen loads. 


UlTTREKSEL 

BESTUIWING VAN MIMETES HIRTUS (PROTEACEAE) DEUR KAAPSE- 
SUIKERBEKKIES EN JANGROENTJIES 

Promerops cafer (Kaapse-suikerbekkie), is ’n belangrike stuifmeeldraer vir Mi- 
metes hirtus omdat dit met kop en keel teen die stuifmeelhokke en -stempels skuur 
wanneer dit nektar soek. P. cafer word deur die gedeeltelike oop bloeiwyses gelok, 
veral as gevolg van die groot hoeveelhede nektar beskikbaar. Maksimum stuifmeel- 
beskikbaarheid en kiemkrag val saam met meeste besoeke deur P. cafer, en lei tot 'n 
betreklike hoe graad van saadvorming. Nectarinia violacea (Jangroentjie) besoek 
M. hirtus af en toe maar dra nie veel stuifmeel oor nie. 


Introduction 

Proteaceae constitute one of the major components of fynbos vegetation 
in the south-western Cape region of South Africa (Taylor, 1978). Protea is 
the only genus within this group whose pollination biology has been studied 
intensively (Rourke and Wiens, 1977; Wiens, Rourke, Casper, Rickart, La 
Pine and Peterson, 1982). Rodents are the pollinators for many dwarf 
species, although birds such as the Cape Sugarbird ( Promerops cafer) and a 
variety of sunbirds appear to be the major pollen vectors in tall, shrubby 


* Permanent address: School of Biology, Western Australian Institute of Techr 
ogy, Kent St., Bentley 6102, Western Australia. 

Accepted for publication 11th November, 1982. 


125 


126 


Journal of South African Botany 


species (Broekhuysen, 1959; Skead, 1967; Mostert, Siegfried and Louw, 
1980). 

Another genus of the Proteaceae that is endemic to the southern and 
south-western Cape, is Mimetes. This comprises twelve species, and occurs 
sporadically from Clanwilliam through the Cape Peninsula and Hottentots 
Holland Mountains, to Formosa Peak in the east (Rourke, 1982). One of 
the most geographically restricted species in this group is Mimetes hirtus , oc- 
curring in isolated populations on the Cape Peninsula, and on the south 
coast near Betty’s Bay, Hangklip and Hermanus (Rourke, 1982). 

Apart from the general observation that M. hirtus inflorescences are visi- 
ted frequently by P. cafer and sunbirds, virtually nothing is known of the 
pollination biology of this species (Vogel, 1954). The main objectives of the 
present paper, therefore, are to establish whether P. cafer and/or sunbirds 
are significant vectors for M. hirtus pollen and whether the selection of 
inflorescences visited is related to the availability of nectar and viable pol- 
len. 


Material and Methods 
Study areas 

Field investigations were conducted at Betty’s Bay, 90 km east of Cape 
Town on the south coast, and at Kirstenbosch National Botanic Gardens in 
Cape Town, during July and August, 1982. Topography, soil structure and 
general vegetation for both of these areas have been described previously 
(Fagan, 1973; Boucher, 1978). 

The study area at Betty’s Bay comprised several stands of Mimetes hirtus 
(Proteaceae) and Erica perspicua (Ericaceae), in which Promerops cafer and 
Orange-breasted Sunbirds ( Nectarinia violacea) were the dominant nectari- 
vores. Male and female P. cafer actively defended non-breeding territories 
in M. hirtus against conspecific birds and TV. violacea. TV. violacea was more 
frequently seen amongst E. perspicua than M. hirtus. The area in which 
plant and bird observations were made was located on a narrow belt of 
coastal seepage fynbos at the southern foot of Elephant Rock Mountain 
(Boucher, 1978). 

Studies at Kirstenbosch were conducted using a group of 13 M. hirtus 
that had been grown from cuttings planted in poorly-drained soil on the low- 
er south-eastern slopes of Table Mountain. No nectarivores were seen visit- 
ing inflorescences during the study period, possibly because of the small 
M. hirtus population, and the availability of numerous alternative nectar 
sources. Nevertheless, this site proved ideal for extended studies of flower- 
ing phenology and diurnal variation in nectar availability. 


Pollination of Mimetes hirtus by sugarbirds and sunbirds 


127 


The Plant 

M. hirtus is a relatively dense shrub, approximately 1,5 m in height, that 
bears numerous terminal flowering heads. Individual heads comprise several 
red and yellow inflorescences (capitula), each of which consists of 8-11 
florets surrounded by involucral bracts. As each floret develops, a bent red 
style emerges from between the bracts, ultimately becoming fully erect when 
the pollen presenter dehisces from the anthers (Figure 1). Four tiny nec- 
taries (hypogynous scales) are located at the base of each inflorescence, ap- 
proximately 57 mm from the distal ends of the fully extended styles 
(Rourke, 1982; Collins, unpublished data). 

In the present study, inflorescences in which only some of the styles were 
fully extended have been categorised as being “partially-open”. Inflores- 
cences with all styles erect, but not yet withered, have been classified as 
“fully-open”. At Betty’s Bay and Kirstenbosch, the numbers of partially- 
and fully-opened inflorescences were counted on randomly-selected flower- 
ing heads of M. hirtus during the study period. 

The Birds 

Male P. cafer weighed 38,2 ± 3,1 g, and had bill lengths of 29,1 ± 1,0 
mm (N = 6), whereas corresponding values for N. violacea were 9,9 ± 0,6 g 
and 21,8 ± 0,9 mm (N = 6), respectively. Female birds were slightly smaller 
than males (Collins, unpublished data). 

Patterns of floret opening 

At Kirstenbosch, 12 marked inflorescences were observed over a period 
of eight days. During this time, florets with erect styles were counted at 
07h30, 10h30, 13h30 and 16h30. Emergent styles were marked with a 
coloured felt pen to facilitate counting and subsequent identification. Data 
gathered were used to estimate diurnal changes in rates at which florets 
opened, and the time which elapsed between emergence of the first and last 
florets on each inflorescence. 


Retention of pollen on pollen presenters 

Pollen smears were taken from pollen presenters at known times after 
anthesis during experiments at Kirstenbosch. The technique used entailed 
rubbing a small cube of agar, impregnated with basic fuchsin, against all 
sides of each pollen presenter. Permanent mounts of cubes were examined 
microscopically, and the numbers of stained pollen grains counted in four 
randomly chosen 320 x fields of view (Collins and Briffa, 1982). The pres 
ence of pollen was categorised as: + + + (> 60 pollen grains per 320 x field 


pollen presenters 



Fig. 1. 

Inflorescence of Mimetes hirtus, showing florets at various stages of development; (a) unopened, (b) partially-opened, 

(c) fully-opened. 


Pollination of Mimetes hirtus by sugarbirds and sunbirds 129 

of view); ++ (10-60 grains per field); + (1-9 grains per field); and 0 (no 
pollen grains visible). 

Additional evidence regarding temporal changes in pollen abundance on 
pollen presenters of M. hirtus was obtained by taking photo-micrographs of 
fresh terminal portions of styles at 0 and 2 days after anthesis, using a Leitz 
Laborlux II compound microscope and Wild Photoautomat MPS 45 unit. 

Pollen viability 

Several styles were collected immediately following anthesis and stored 
in clean, dry, sealed glass vials. Three pollen samples taken from these 
styles were smeared onto sterile 1,5 % tap water nutrient agar (Oxoid CM3) 
at 0, 2 and 5 days after collection. Each smear was examined daily for 5 
days, and the percentage germination of pollen grains determined (Stanley 
and Linskens, 1974). Pollen viability was also tested for at 0 and 7 days by 
placing samples in drops of 4 % sulphuric acid and looking for signs of “ger- 
mination” (Stanley and Linskens, 1974). 

Nectar availability 

Nectar standing crop volumes and percentage equivalent sucrose concen- 
trations (w/w) were measured for individual inflorescences at 07h30, 10h30, 
16h30 and 13h30 or 21h30 on two days at Betty’s Bay, and four days at 
Kirstenbosch. On most occasions, five closed, five partially-opened and 10 
fully-opened inflorescences were sampled. Volumes were measured using 
Drummond calibrated capillary tubes, and concentrations with a National 
temperature-compensated refractometer (Collins and Briffa, 1982; 1983). 

Nectariv ore foraging behaviour 

At Betty’s Bay, the manner in which P. cafer and N. violacea probed 
M. hirtus inflorescences was recorded at various times of day. Where poss- 
ible, it was also noted whether the inflorescences visited were closed, par- 
tially-open or fully-open. No observations of this kind were possible at 
Kirstenbosch. 

Pollen transfer by nectarivores 

The relative importance of P. cafer and N. violacea as vectors of M. hir- 
tus pollen, and the manner in which pollen transfer was effected, were as- 
sessed by examining smears taken from the foreheads, beaks and throats of 
birds caught in mist nets at Betty’s Bay. Smears were prepared in essentially 
the same way as described earlier, and pollen grains counted after identifica- 
tion by comparison with those in smears taken from known plants (Collins 
and Briffa, 1982). 


130 


Journal of South African Botany 


Effectiveness of pollination and seed set 

Ten flowering heads of M. hirtus at Betty’s Bay were enclosed with 
10 mm mesh chicken wire to exclude P. cafer and N. violacea. Selection of 
flowering heads was such that none bore partially- or fully-opened inflores- 
cences at the time of enclosure (August). In November, these heads were 
examined for evidence of seed set, as were another 10 heads that had not 
been screened from nectarivorous birds. 

Results 

Floret opening 

Rates at which florets on marked Mimetes hirtus inflorescences at 
Kirstenbosch produced erect styles are outlined in Table 1. There was a con- 
sistent trend over four days for highest rates to occur between 10h30 and 
13h30, and lowest during the “overnight” period (Kendall’s W = 0,78, x 2 = 
9,30, P < 0,01). Only 12 of the marked inflorescences changed from being 
unopen to completely open during the study period. Of these, two opened 
in one to two days, two took almost eight days, and the remainder opened 
at fairly regular intervals between these times (x ± SD = 4,3 ±2,1 days). 


Table 1. 

Opening of M. hirtus florets on marked inflorescences at Kirstenbosch. 


Date and parameter 

Time of day 


16h30-07h30 

(“overnight”) 

07h30-10h30 

10h30-13h30 

13h30-16h30 

9-10 August 





Inflor 

20 

16 

17 

17 

Florets opening 

27 

7 

16 

5 

Florets per inflor. h . . . . 

0,09 

0,15 

0,31 

0,10 

10-11 August 





Inflor 

17 

15 

17 

17 

Florets opening 

7 

0 

15 

7 

Florets per inflor. h . . . . 

0,03 

0,00 

0,29 

0,14 

11-12 August 





Inflor 

16 

16 

16 

15 

Florets opening 

14 

4 

21 

6 

Florets per inflor. h . . . . 

0,06 

0,08 

0,44 

0,13 

12-13 August 





Inflor 

13 

13 

13 

11 

Florets opening 

0 

8 

15 

1 

Florets per inflor. h . . . . 

0.00 

0,21 

0,38 

0,03 


Pollination of Mimetes hirtus by sugarbirds and sunbirds 


131 


The numbers of fully- and partially-opened inflorescences on 12 marked 
M. hirtus flowering heads at Kirstenbosch varied slightly during the study 
period, although values of 3,4 ± 2,3 and 1,7 ± 1,2 inflorescences per head 
respectively, on 9th August give a typical indication of relative abundance. 
Corresponding values for 39 marked flowering heads at Betty’s Bay on 25th 
July were 4,5 ± 3,5 and 1,2 ± 1,7 inflorescences per head. Comparisons of 
equivalent values for the two sites show no significant differences, although 
there were always more fully- than partially-opened inflorescences present 
(Kirstenbosch: t = 2,27, P < 0,05; Betty’s Bay: t=5,30, P < 0,001). 




r iu . \ t 

Terminal portions of Mimetes hirtus styles, showing pollen on pollen presenters (a) at 
time of anthesis, and (b) 2 days later. Magnification is 40x natural size. 


132 


Journal of South African Botany 


Retention of pollen on florets 

Pollen is extremely abundant on freshly exposed pollen presenters of 
M. hirtus (Figure 2, Table 2). Exposure for 15 to 18 hours, however, re- 
duced abundance significantly. After two to three days, there was virtually 
no pollen left on presenters at Kirstenbosch. 

Pollen viability 

Germination of freshly-released pollen on nutrient agar was approxi- 
mately 30 % successful (Table 3). However, viability of pollen decreased sig- 
nificantly with age, with five-day-old pollen exhibiting only 3 % germination. 

Table 2. 

Presence of pollen grains in smears taken from pollen presenters of M. hirtus. 


Time after exposure of pollen presenter (h) 


0 

0-3 

3-6 

15-18 

>192 

+ + + 

+ + + 

+ + + 

+ + 

+ 

+ + + 

+ + + 

+ + + 

+ + 

+ 

+ + + 

+ + + 

+ + + 

+ + 

+ 

+ + + 

+ + 

+ + 

+ 

+ 

+ + + 

+ + 

+ + 

+ 

0 

+ + + 

+ + 

+ + 

+ 

0 

+ + + 

+ + 

+ + 

+ 

0 

+ + + 

+ + 

+ + 

+ 

0 


Five 320 x microscopic fields of view were examined on each of 40 smears; + + + 
denotes an average of >60 pollen grains per field of a particular smear; ++ 10-60 
grains per field; + 1-9 grains per field; 0 no grains visible. 


Table 3. 

Germination of M. hirtus pollen grains at intervals following exposure on pollen 

presenters. 


Date trial 
initiated 

% germination 
on day 1 

% germination 
on day 3 

% germination 
on day 6 

17 August 

0,0 ± 0,0 
(300) 

33,3 ± 2,7 
(186) 

28,9 ± 3,8 
(187) 

19 August 

0,0 ± 0,0 
(300) 

16,7 ± 10,9 
(180) 


22 August 

0,0 ± 0,0 
(300) 

3,2 ± 2,1 
(181) 



Numbers in parentheses denote total pollen grains counted on each occasion (not 
all grains on agar plates counted). 


Pollination of Mimetes hirtus by sugarbirds and sunbirds 


133 


A similar pattern was shown by data generated when fresh or seven-day-old 
pollen was placed in 4 % sulphuric acid. “Germination” was 84,7 ± 3,6 % 
successful in the first instance, and 4,8 ± 4,3 % in the latter (at least 200 pol- 
len grains counted in each of the three samples for both treatments; t = 
24,68; P < 0,001). 

Nectar availability’ 

Nectar standing crop volumes for partially- and fully-opened M. hirtus 
inflorescences at Betty’s Bay are shown in Table 4. Values do not vary signi- 
ficantly from one time of day to another or between fully- and partially- 
opened inflorescences, except for volumes at 07h30 in partially-opened 
inflorescences, which are significantly greater than most other values 
(P < 0,01 for 5 out 7 possible comparisons on 28th August). 

At Kirstenbosch, standing crop volumes for unopened, partially- and 
fully-opened inflorescences were much greater than those at Betty’s Bay, 
and showed no significant diurnal variations (Figure 3). Volumes were gen- 
erally similar for the same types of inflorescences, at the same times of day, 
despite changes in air temperature and incidence of rain. Nevertheless, 
there was a consistent trend for smaller volumes to occur in fully- rather 
than partially-opened inflorescences (Mann-Whitney U > 42, P < 0,05, in 9 
out of 14 possible comparisons; t > 2,36, P < 0,05, in 12 of the same com- 
parisons). Volumes in unopened inflorescences were variable, usually being 
very small except in instances where the inflorescences opened readily when 
touched. 


Table 4. 

Nectar standing crop volumes (/x€) for M. hirtus at Betty’s Bay. 


Date and 
condition 
of inflor. 

Time of day 

07h30 

10h30 

13h30 

16h30 

21h30 

26 July (a) 

(b) 

28 Aug. (a) 

(b) 






3,3 ± 8,4 
22,6 ± 21,9 
2,8 ± 5,3 

0,8 ± 1,3 
3,8 ± 3,4 
4,1 ± 9,1 

0,9 ± 1,0 
1,6 ± 2,6 

4.1 ± 12,6 

2.1 ± 2,2 
0,8 ± 2,0 

4,1 ± 7,6 


Sample size was 12 in every instance; (a) and (b) denote partially- and fully-open 
ed inflorescences, respectively. 


nectar standing crop {p.t per inflorescence) 


134 Journal of South African Botany 



UPO UPO UPO UPO 

07h30 10h30 13h30 16h30 

time of day 

Fig. 3. 


30 

20 

10 

0 

30 

20 

10 

0 

30 

20 

10 

0 

30 

20 

10 

0 


environmental temperature (°C) 


Pollination of Mimetes hirtus by sugarbirds and sunbirds 


135 


Additional measurements at Kirstenbosch showed that standing crop vol- 
umes varied according to the number of erect styles in partially-opened 
inflorescences (323,1 ± 172,1 /x€ when one to two styles erect, 183,6 ± 
93,4 /x6 when seven or more styles erect; N = 10 in each instance; t = 2,25; 
P < 0,05). Data used for this comparison were collected at 10h30, three 
days after the main experimental period, although overall volumes for the 
20 inflorescences (253,4 ± 152,6 /x6) did not differ significantly from those 
indicated in Figure 3. 

Few nectar concentrations were measured at Betty’s Bay, but the values 
obtained (16,1 ± 1,6%, N = 5) were similar to those for Kirstenbosch 
(Table 5). There was a tendency for nectar concentrations at Kirstenbosch 
to be slightly greater for fully- than partially-opened inflorescences, although 
differences were not statistically significant. 

Bird foraging behaviour 

Without exception, Promerops cafer foraged on M.. hirtus inflorescences 
at Betty’s Bay while perching on top of the flowering heads (150 obser- 
vations). Consequently, the birds had to reach over the distal ends of any 
erect styles in order to probe the inflorescences, thereby brushing their 
throats against the stigmas and pollen presenters. Fewer Nectarinia violacea 
were seen foraging on M. hirtus (20 observations), although all of these con- 
sistently perched to the side of, or beneath, the inflorescences that they 
probed. As a consequence, N. violacea rarely brushed against the stigmas or 
pollen presenters of erect styles. 

It was not possible to watch individual N. violacea for long enough at 
Betty’s Bay to ascertain whether they preferred M. hirtus inflorescences at 
particular stages of development. However, a territorial female P. cafer that 
could be watched for most of the day showed a clear preference for partial- 
ly-opened inflorescences (unopened: 107 probes; partially-opened: 704 
probes; fully-opened: 113 probes). It is not known whether this pattern is 
typical of P. cafer, in general. Nevertheless, the observed total numbers of 
inflorescences probed per flowering head by several P. cafer and N. violacea 
were significantly greater than the numbers of partially-opened inflores- 


Fic. 3. 

Nectar standing crop volumes for Mimetes hirtus inflorescences at Kirstenbosch. 
Vertical lines indicate ranges of values, horizontal lines means, and rectangles one 
standard deviation either side of the means. The letters U (sample size 5), P (sample 
size 5) and 0 (sample size 10) represent data for unopened, partially- and fully-open- 
ed inflorescences, respectively. Figures enclosed in circles denote the numbers o 
closed inflorescences that opened partly when touched. Symbols • and • ' ' denote 
air temperatures and occurrence of rain, respectively, at various times of day. 


136 


Journal of South African Botany 


Table 5. 

Nectar concentrations (% w/w) for M. hirtus at Kirstenbosch. 


Date and 
condition 
of inflor. 


Time of day 



07h30 

10h30 

13h30 

16h30 

9 Aug. (a) 



17,7 ± 2,2 
(6) 

17,4 ± 0,6 
(5) 

(b) 



16,7 ± 3,1 
(5) 

21,0 ± 4,3 
(5) 

10 Aug. (a) 

16,7 ± 2,8 

(5) 

16,5 ± 1,9 
(5) 

14,6 ± 1,2 
(4) 

16,5 ± 2,5 

(4) 

(b) 

18,8 ± 2,6 
(?) 

16,6 ± 1,9 
(3) 

15,8 ± 1,0 
(4) 

14,8 ± 4,5 
(5) 

11 Aug. (a) 

13,5 ± 1,2 

(5) 

15,7 ± 1,1 
(5) 

15,5 ± 1,8 
(5) 

14,5 ± 0,9 
(5) 

(b) 

18,0 ± 5,9 
(6) 

16,0 ± 1,7 
(5) 

16,2 ± 2,1 
(3) 

16,5 ± 2,4 

(4) 

12 Aug. (a) 

16,0 ± 1,3 
(5) 

15,4 ± 1,2 
(5) 

15,4 ± 0,7 

(5) 

14,4 ± 0,7 

(5) 

(b) 

16,8 ± 1,2 
(5) 

19,6 ± 3,8 
(4) 

20,3 ± 5,5 
(2) 

18,8 

(1) 


(a) and (b) have same meanings as in Table 4. Sample sizes given in parentheses. 


Table 6. 

Comparison of observed abundances of inflorescences and nectarivore probing on 
flowering heads of M. hirtus. 



Partially- 
opened inflor. 
per head 

Fully-opened 
inflor. per 
head 

Inflor. probed 
per head by 
6 P. cafer 

Inflor. probed 
per head by 
2 N. violacea 

X 

1,2 

4,5 

3,6 

3,0 

SD 

1,7 

3,5 

2,4 

3,0 

Range 

0-7 

0-14 

1-11 

1-5 

N 

39 

39 

67 

28 


Pollination of Mimetes hirtus by sugarbirds and sunbirds 


137 



138 


Journal of South African Botany 


cences available per head at any given time (Table 6; P. cafer: t = 5,49, 
P < 0,001; N. violacea : t = 3,12, P < 0,01) 

Pollen transfer 

The majority of pollen grains collected from P. cafer and N. violacea at 
Betty’s Bay were derived from M. hirtus (Table 7). Variability of pollen 
grain counts in different smears was considerable, although it is obvious that 
N. violacea carried significantly less pollen than P. cafer. Similar pollen 
loads were borne by head and throat regions, regardless of the nectarivore 
species involved. 

Effective seed set 

Seed set in 10 M. hirtus flowering heads to which nectarivores had access 
at Betty’s Bay was relatively high, and significantly greater than in 10 heads 
enclosed by wire mesh (Table 8). 


Table 8. 

Seed set in M. hirtus at Betty’s Bay. 



Exposed flowering 

Enclosed flowering 

Parameter 

heads 

heads 

Total inflorescences studied 

107 

121 

Total inflorescences containing seed .... 

38 

1 

% seed set 

35,5 

0,8 


Discussion 

There has been a resurgence of interest in the pollination biology of 
plants in recent years (e.g. Faegri and van der Pijl, 1971; Heinrich, 1975; 
Wolf, 1975; Rourke and Wiens, 1977; Carpenter, 1978; Armstrong, 1979; 
Ford, Paton and Forde, 1979; Hopper, 1980; Frost and Frost, 1981; Collins 
and Briffa, 1982; Collins and Briffa, 1983). Much has been written regarding 
pollination and utilization of floral nectar by mammals, birds and inverte- 
brates, although attention has also been paid to the significance of pollen 
transfer in terms of gene flow and plant population structure (e.g. Beattie, 
1978; Levin, 1978). 

Despite the general interest shown in pollination in other parts of the 
world, relatively little attention has been paid to this aspect of plant biology 
in southern Africa. Virtually the only major studies in this area have con- 
cerned the pollination of various Protea species (Rourke and Wiens, 1977; 


Pollination of Mimetes hirtus by sugarbirds and sunbirds 


139 


Wiens et al., 1982), and Strelitzia nicolai (Frost and Frost, 1981). In the lat- 
ter case, a variety of sunbirds (Nectarinidae) appear to be the major pollen 
vectors and users of floral nectar. Some Protea species are also pollinated by 
sunbirds and Cape Sugarbirds, although rodents appear to transfer pollen in 
many instances. Invertebrates do not seem to play a vital role for any of 
these plants, although they may do so in other groups. 

It has been suggested that members of the Proteaceae such as Mimetes 
are pollinated primarily by birds (Skead, 1967; Rourke, 1982). Qualitative 
data have not been published previously in support of this contention, 
although it is obvious that inflorescences of members of this genus have 
many features that are allegedly characteristic of bird-pollinated plants: 
bright colouration, ease of access from above, diurnal availability of nectar 
and tubular perianths (Faegri and van der Pijl, 1971; Carpenter, 1978; Hop- 
per, 1980). Direct observations made in the present study, however, confirm 
that, at least at Betty’s Bay, Promerops cafer and Nectarinia violacea are fre- 
quent visitors to Mimetes hirtus. Intermittent observations of inflorescences' 
during the day and at night failed to detect any mammalian visitors. The 
only invertebrates present were small Hemiptera (2,7 ± 2,5 per inflores- 
cence, N = 10), and these appeared to be confined to perianth tubes, rarely 
moving on to the styles (Collins, unpublished data). 

Evidence obtained at Betty’s Bay during the present study demonstrates 
that P. cafer is a major pollen vector for M. hirtus , although N. violacea ap- 
pears to be much less important. This difference could be a function of the 
strongly territorial behaviour of P. cafer, which allows N. violacea relatively 
few opportunities to probe inflorescences. It is likely, however, that differ- 
ences in foraging behaviour are more important. As the subordinate nectari- 
vore, N. violacea is probably forced to rely on fully-opened inflorescences, 
with their relatively low nectar rewards. Consequently it would only rub its 
head and throat against dehisced anthers. The existence of substantial pollen 
loads on both throats and heads of P. cafer is consistent with the foraging 
behaviour of this species, particularly as its preference is for partially-opened 
inflorescences in which it could presumably induce anthesis as its head 
brushed against unopened florets. The significance, if any, of higher rates of 
floret opening between 10h30 and 13h30 is not apparent. 

M. hirtus is protandrous, with pollen exposure on a given floret occurring 
before the stigmatal groove is receptive (Rourke, personal communication). 
Effective transfer of pollen between one floret and another is therefore de- 
pendent upon vectors such as P. cafer brushing against pollen presenters and 
stigmata at appropriate stages in their development. Pollen availability c r 
individual presenters decreases significantly within 15 h of anthesis, even 
at sites such as Kirstenbosch where nectarivores rarely visit M. hirtus. How- 
ever, it takes several days for all florets on an inflorescence to open, thereby 


140 


Journal of South African Botany 


extending the duration of pollen-availability on the inflorescence. Neverthe- 
less, viability of any pollen that remains on pollen presenters is reduced sig- 
nificantly after two days. It is obvious, therefore, that effective outcrossing is 
unlikely unless vectors visit inflorescences that are partially- or just com- 
pletely-opened. In this context, it would seem significant that P. cafer has a 
definite preference for partially-opened inflorescences, although it some- 
times probes others. It is not known when the stigmatal grooves of 
M. hirtus are receptive, although this would most appropriately occur when 
birds are visiting inflorescences on which most of the florets were open. 

Partially-opened M. hirtus inflorescences at Kirstenbosch produce stand- 
ing crop volumes of relatively dilute nectar that are larger than most values 
reported for other plants (e.g. Ford, 1979; Frost and Frost, 1981; Collins 
and Briffa, 1982; 1983). There is no significant variation from one time of 
day to another, possibly because nectarivores rarely visit the inflorescences. 
With few exceptions, unopened and fully-opened inflorescences contain 
little nectar. These observations suggest that nectar production by M. hirtus 
varies according to the age of the inflorescences, a trend similar to that re- 
ported for several other plant species (e.g. Frost and Frost, 1981). Standing 
crops at Betty’s Bay are much smaller than at Kirstenbosch, possibly be- 
cause of differences in soil water potential (Fagan, 1973; Boucher, 1978) and 
greater nectar demand by P. cafer and N. violacea. At this site, nectarivore 
pressure during the day is almost certainly responsible for the lack of signi- 
ficant differences between standing crops for partially- and fully-opened 
inflorescences at all times except shortly after dawn (07h30). This explana- 
tion is consistent with the observed foraging preference of P. cafer for par- 
tially-opened inflorescences. 

The M. hirtus-P. cafer system is clearly well adapted for pollen transfer. 
Nectar rewards are greatest in inflorescences that are releasing pollen, there- 
by attracting P. cafer at a time when maximum pollen transfer can occur. 
The effectiveness of this system is demonstrated by the relatively high level 
of seed set that results from exposure of M. hirtus to P. cafer (and N. viola- 
cea). Restriction of M. hirtus to small isolated populations in the south- 
west Cape would not seem to be caused by any deficiencies in this system, 
but rather because of the need for habitats containing acidic, peaty soils 
(Rourke, personal communication) and/or the territorial nature of the major 
pollinator, P. cafer. 


Acknow ledgements 

Funds provided by the Fynbos Biome Project, Jack Niven Foundation 
and University of Cape Town supported this study, while the author was on 
sabbatical leave from the Western Australian Institute of Technology. Pro- 


Pollination of Mimetes hirtus by sugarbirds and sunbirds 


141 


fessor H. B. Rycroft and Mr. J. Winter kindly gave permission for some of 
the work to be conducted at Kirstenbosch National Botanic Gardens. Mr. 
R. Attwell and other councillors gave permission for field studies to be con- 
ducted within the Municipality of Betty’s Bay. Valuable assistance with field 
work was given by Mrs. L. Collins and Mr. R. Collins. Mr. D. Gerneke 
helped with photo-micrography, and Professor D. Mitchell provided facili- 
ties and materials required for culturing pollen grains. Dr. J. Rourke gave 
valuable guidance throughout the study, and helped arrange accommodation 
at the Harold Porter Lodge at Betty’s Bay. Dr. Rourke and Professor W. R. 
Siegfried made constructive comments on a draft of the manuscript. 


References 

Armstrong, J. A., 1979. Biotic pollination mechanisms in the Australian flora — a re- 
view. N.Z. Jnl. Bot. 17: 467-508. 

Beattie, A., 1978. Plant-animal interactions affecting gene flow in Viola. In: Rich- 
ards, A. J. (ed.), The Pollination of Flowers by Insects, pp. 151-164. Lon- 
don: Academic Press. 

Boucher, C., 1978. Cape Hangklip area. II. The vegetation. Bothalia 12: 455-497. 

Broekhuysen, G. J., 1959. The biology of the Cape Sugarbird Promerops cafer. Os- 
trich Suppl. 3: 180-221. 

Carpenter, F. L., 1978: Hooks for mammal pollination? Oecologia 35: 123-132. 

Collins, B. G. and Briffa, P., 1982. Seasonal variations in abundance and foraging 
of three species of Australian honeyeaters. Aust. Wild l. Res. 9: 557-569. 

Collins, B. G. and Briffa, P., 1983. Nectar utilisation and pollination by Australian 
honeyeaters and insects visiting Calothamnus quadrifidus (Myrtaceae). 
Oecologia (submitted). 

Faegri, K. and van der Pijl, L. 1971. Principles of Pollination Ecology. New York: 
Pergamon Press. 

Fagan, G., 1973. Kirstenbosch Development Study. Report prepared for Kirsten- 
bosch Botanic Garden Board. 

Ford, H. A., 1979. Interspecific competition in Australian honeyeaters-depletion of 
common resources. Aust. J. Ecol. 4: 145-164. 

Ford, H. A., Paton, D. C. and Forde, N., 1979. Birds as pollinators of Australian 
plants. N.Z. Jnl. Bot. 17: 509-519. 

Frost, S. K. and Frost, P. G. H., 1981. Sunbird pollination of Strelitzia nicolai. Oe- 
cologia (Berl.) 49: 379-384. 

Heinrich, B., 1975. Energetics of pollination. Annu. Rev. Ecol. & Syst. 6: 139-170. 

Hopper, S. D., 1980. Bird and mammal pollen vectors in Banksia communities at 
Cheyne Beach, Western Australia. Aust. J. Bot. 28: 61-75. 

Levin, D. A., 1978. Pollinator behaviour and the breeding structure of plant popu- 
lations. In: Richards, A. J. (ed.), The Pollination of Flowers by Insects, pp 
133-150. London: Academic Press. 

Mostert, D. P., Siegfried, W. R. and Louw, G. N., 1980. Protea nectar and satel- 
lite fauna in relation to the food requirements and pollinating role ot the 
Cape Sugarbird. S. Afr. J. Sci. 76: 409-412. 


142 


Journal of South African Botany 


Rourke, J. P., 1982. Mimetes. Cape Town: Tiyan Publishers. 

Rourke, J. and Wiens, D., 1977. Convergent floral evolution in South African and 
Australian Proteaceae and its possible bearing on pollination by non-flying 
mammals. Ann. Mo. bot. Gdn. 64: 1-17. 

Skead, C. J., 1967. The Sunbirds of Southern Africa. Cape Town: Balkema. 

Stanley, R. G. and Linskens, H. F., 1974. Pollen. Berlin: Springer-Verlag. 

Taylor, H. C., 1978. Capensis. In: Werger, M.J.A. (ed.), Biogeography and Ecol- 
ogy of Southern Africa, pp. 171-229. The Hague: Junk. 

Vogel, S., 1954. Bliitenbiologische Typen Als Elemente Der Sippengliederung. Jena: 
Fischer-Verlag. 

Wiens, D., Rourke, J. P., Casper, B. B., Rickart, E. A., LaPine, T. R. and Peter- 
son, C. J., 1982. Non-flying mammal pollination of southern African pro- 
teas: a non-coevolved system. Ann. Mo. bot. Gdn. (in press). 

Wolf, L. L., 1975. Energy intake and expenditure in a nectar feeding sunbird. Ecol- 
ogy 56: 92-104. 


J1 S. Afr. Bot. 49 (2): 143-159 (1983) 


NEW XANTHOPARMELIAE (LICHENES) FROM SOUTHERN AND 
CENTRAL AFRICA 

M. D. E. Knox 

(Department of Botany, University of the Witwatersrand, 1 Jan Smuts Ave- 
nue, Johannesburg 2001, R.S.A.) 

F. A. Brusse 

(Botanical Research Institute, Private Bag X101, Pretoria 0001, R.S.A.) 
Abstract 

Ten new species of Xanthoparmelia are described from southern and central 
Africa, and one new combination in the genus is proposed. The new species are X. 
aggregata Knox, X. cedrus-montana Brusse, X. dysprosa Brusse et Knox, X. ianthina 
Brusse, X. karoo Knox et Brusse and X. leucostigma Brusse from the Karoo and 
Western Cape, X. cylindriloba Knox and X. hedbergii Knox from Kenya, X. fucina 
Knox from Zimbabwe and X. walteri Knox from South West Africa/Namibia. The 
new combination is X. exornata (Zahlbr.) Brusse et Knox from the Karoo. 

UlTTREKSEL 

NUWE XANTHOPARMELIAE (LICHENES) VAN SUIDELIKE EN SEN- 
TRAAL AFRIKA 

Tien nuwe soorte van Xanthoparmelia van suidelike en sentraal Afrika word be- 
skryf, en een nuwe kombinasie in die genus word voorgestel. Die nuwe soorte is X. 
aggregata Knox, X. cedrus-montana Brusse, X. dysprosa Brusse et Knox, X. ianthina 
Brusse, X. karoo Knox et Brusse en X. leucostigma Brusse van die Karoo en Wes- 
telike Kaap, X. cylindriloba Knox en X. hedbergii Knox van Kenia, X. fucina Knox 
van Zimbabwe en X. walteri Knox van Suidwes Afrika/Namibia. Die nuwe kombina- 
sie is X. exornata (Zahlbr). Brusse et Knox van die Karoo. 

Introduction 

Lichens now placed in the genus Xanthoparmelia were first recognised as 
a distinct group by Nylander (1860), who placed them in his “Stirps Parme- 
liae conspersae”. The distinction was further consolidated by Wainio (1890), 
who coined the name Xanthoparmelia for this section of the subgenus Eu- 
parmelia Nyl., placing in the section all narrow-lobed yellow-green species 
(containing the pigment usnic acid in the upper cortex). In elevating the 
group to subgeneric status Hale, in Hale and Kurokawa (1964), employed 
substrate ecology, rhizine morphology and the absence of cilia to refine ,Wa> 
nio’s (1890) concept, and this interpretation was retained when Hale (1974) 


Accepted for publication 25th October, 1982. 


143 


144 


Journal of South African Botany 


raised the group to generic status. This final step has been rejected by Belt- 
man (1978), Krog and Swinscow (1979), Hawksworth et al. (1980), Elix 
(1981) and Poelt and Vezda (1981), although Krog and Swinscow (1979) and 
Elix (1981) accept the subgeneric ranking. Elix (1981) pointed out that in 
the genera Xanthoparmelia, Pseudoparmelia and Neofuscelia, all genera 
which have been segregated from Parmelia sensu lato, striking morpholo- 
gical parallels may be found, the generic segregation in such cases being ef- 
fectively based solely on the chemical constituents present in the upper cor- 
tex. ThiS may be argued to be inadequate grounds, and certainly suggests 
that additional criteria must be sought and considered in an attempt to re- 
solve this problem. 

In the present paper, no attempt is made further to consider this ques- 
tion, but Xanthoparmelia is recognised at generic level because of the dis- 
tinctive morphology of the majority of species, because of its ability to sur- 
vive and grow in arid conditions, and because of its characteristic tendency, 
by comparison with the other two genera mentioned above, to produce lux- 
uriant growth in moister environments. 

The following species have been described from specimens collected 
from the Karoo and adjacent regions by one author (F.A.B.) during the 
course of a taxonomic investigation into the genus presented for the degree 
of M.Sc. at the University of the Witwatersrand, and from specimens seen 
by the other author at the Smithsonian Institution, Washington, D.C., while 
on extended study leave from the same University. 

Xanthoparmelia aggregata Knox, sp. nov. Fig. 1. 

Thallus sordide flavescenti-virescens, stellatus, usque ad 50 mm di- 
ametro, terricola, laxe affixus. Lobi 0,5-1, 5 mm lati, ligulati vel canaliculati, 
ramosi, abrupte decrescentibus apicem versus. Pagina infera concolora vel 
quam pagina superior parum fuscans, maculae aliquot atrantes, cum rhizinis 
plerumque consociatae, exclusae. Isidiis sorediisque carens. Cortex superior 
40-50 pm crassus, stratum gonidiale 60-65 pm crassum, sed stratum goni- 
diale 18-20 pm crassum cum pagina infera concolora consociatum, medulla 
alba 220-240 pm crassa, cortex inferior 18-20 pm crassus. Apothecia pycni- 
diaque ignota, sed perithecia parasitica adsunt. 

Cortex superior K non reagens (acidum usnicum continens), medulla 
alba K straminea; C, KC pallide rosea, P non reagens (acidum evernicum 
continens). 

Typus: CAPE — 3118 (Vanrhynsdorp): Salt River (-BC), van Rhyn’s Dorp Division 
(Cape Province), on ground, Sept., 1941, Stokoe 7721 (BOL, holotype; US, iso- 
type). 

Other Specimen Seen 

N. S. Pillans s.n., January 1928 (BOL). 


New Xanthoparmeliae from southern and central Africa 145 

Xanthoparmelia aggregata is a member of the X. amphixantha (Mull. 
Arg.) Hale morphological group, based on an Australian species, X. am- 
phixantha, containing stictic, norstictic and constictic acids, and including 
Parmelia ( Xanthoparmelia ) pseudoamphixantha Elix (norstictic and con- 
norstictic acids), P. (X.) reptans Kurok. (fumarprotocetraric and succinpro- 
tocetraric acids) and P. (X.) willisii Kurok. (fumarprotocetraric, succinpro- 
tocetraric and protocetraric acids). From South Africa X. subflabellata 
(Steiner) Hale contains stictic, norstictic and constictic acids, and may well 
be synonymous with X. amphixantha. 

All the above lichen acids are 13-orcinol depsidones, which is the most 
common group of such substances in the genus Xanthoparmelia. Xanthopar- 
melia aggregata, however, contains evernic acid, an orcinol p-depside, which 
although not closely related to the 6-orcinol depsidones, is clearly simpler in 
structure, and could therefore be considered to be more primitive. Other or- 
cinol p-depsides found in Xanthoparmelia species are lecanoric acid, found 
in X. worcesteri (Steiner et Zahlbr.) Hale, X. joranadia (Nash) Hale, X. 
arida Egan et Derstine and X. lecanorica (Hale) Hale, and gyrophoric acid, 
found in X. leucostigma Brusse, described later in this paper. Evernic acid is 
found also in X. dysprosa Brusse et Knox, another new species described 
herein. This small proportion, 8 out of approximately 170 species, empha- 
sises the relative rarity of these substances in Xanthoparmelia , but the scat- 
tered distribution — X. joranadia and X. arida in North America and the re- 
mainder in South Africa — together with the considerable diversity of thallus 
morphology in these species, does not permit any simple explanation of 
their occurrence. It appears that orcinol p-depsides must have arisen inde- 
pendently on a number of occasions. 


Xanthoparmelia cedrus-montana Brusse, sp. nov. Fig. 2. 

Thallus foliosus, nitidus vel sordide flavovirens, laxe vel moderate adna- 
tus, saxicola, usque ad 100 mm diametro. Lobi sublineares vel elongati, im- 
bracati, 0, 3-2,0 mm lati, leviter vel valde albomaculati. Pagina infera nigra, 
rhizinis absentibus vel moderate abundantibus. Isidiis sorediisque carens. 
Cortex superior 25-50 pm crassus, stratum gonidiale discontinuum, 0- 
50 pm crassum, medulla alba 100-200 pm crassa, cortex inferior 15-20 ^.m 
crassus. Apothecia cupulis vadosis, substipitata, usque ad 6 mm diametro, 
rara. Hymenium 40-60 pm crassum, subhymenium 10-20 pm crassum, exci- 
pulum 30-60 pm crassum. Sporae 8, 8,0-12,0 x 4, 5-7,0 pm. Pycnidia 
100-150 pm diametro, pycnidiosporae non visae. 

Cortex superior K non reagens (acidum usnicum continens), medulla 
alba K aurea sed armeniacescens, C non reagens, P aurantiaca (acidum 
thamrtolicum continens). 


146 


Journal of South African Botany 


Typus: CAPE — 3319 (Worcester): Ceres (-AD), June 1924, T. B. Leslie 476 
(TUR-V 34575). 

Other Specimen Seen 

Brusse 772-8-2-19, 8/2/77 (J). 

The holotype specimen is annotated, apparently by Leslie, as “var. hy- 
pomelaena Vainio”, referring to Parmelia stenophylla f. hypomelaena Vainio 
ex Lynge (1937), as annotated by Hale. The epithet hypomelaena is, how- 
ever, unavailable in Xanthoparmelia at the specific level, having been used 
by Hale (1974). 

This further addition to the X. hypoleia (Nyl.) Hale complex appears to 
be restricted in distribution to the Cedarberg Range in the Western Cape. A 
further survey carried out in 1981 produced no additional material. 

The X. hypoleia complex includes at present specimens known to contain 
at least 11 different medullary substances, distributed between 9 species, in- 
cluding the three in the present paper. Relationships between these have 
been discussed (Knox, 1981), and will be the subject of a later publication. 



Fig. 1. 

Xanthoparmelia aggregata Knox (holotype). Scale in mm. 


New Xanthopcirmeliae from southern and central Africa 


147 



Fig. 2. 

Xanthoparmelia cedrus-montana Brusse (holotype). Scale in mm. 


Xanthoparmelia cylindriloba Knox, sp. nov. Fig. 3. 

Thallus terricola, pallide flavovirens vel atrans. Lobi 0,5-1, 5 mm lati, 
plus minusve cylindrici post substitutionem inferiori cortici superiore cortice. 
Pagina infera nigra, rhizinis nigris caespitosis. Isidiis sorediisque carens. 
Cortex superior 25-30 /xm crassus, stratum gonidiale 40-50 /im crassum, 
medulla alba 320-520 /xm crassa, cortex inferior 20-25 /xm crassus. Apoth- 
ecia pycnidiaque ignota. 

Cortex superior K non reagens (acidum usnicum continens), medulla 
alba K flavescens turn sanguinea, C non reagens, P aurantiaca (acidum sala- 
zinicum et acidum norsticticum minutissimum continens). 

Typus: KENYA COLONY: Mt. Kenya, Teleki Valley, alpine region, in the upper- 
most part of the valley, on frost-lifting ground, 4 200 m. 27/7/1948, Hedberg s.n., 
Flora Africana 1705 (UPS, holotype; US, isotype). 

Other Collection 

TANGANYIKA TERRITORY: Kilimanjaro, above Peters Hut, 3 970 m 
23/6/1948, Hedberg s.n., Flora Africana 1354 (UPS). 

Xanthoparmelia cylindriloba is a very distinctive species with erect, more 
or less cylindrical lobes, growing on soil. In both the specimens examined. 


148 


Journal of South African Botany 


these lobes were separate, and it may be that in younger specimens a basal, 
broad-lobed region also occurs. No evidence of such lobes could, however, 
be found. 

Although many species of Xanthoparmelia contain salazinic acid, the dis- 
tinctive morphology of X. cylindriloba clearly separates it from other species 
in the genus. 

Xanthoparmelia dysprosa Brusse et Knox, sp. nov. Fig. 4. 

Thallus sordide flavescenti-virescens, saxicola, laxe adnatus vel pulvina- 
tus. Lobi lati turn divaricate ramosi lineares 0,5-1, 5 mm in latitudinem, 
valde albomaculati. Isidiis soredisque carens. Pagina infera nigra, nuda vel 
sparsim rhizinata. Cortex superior 14-17 pm crassus, stratum gonidiale in- 
terruptum, 60-65 pm crassum, medulla alba 120-130 pm crassa, cortex in- 
ferior 16-18 pm crassus. Apothecia dispersa, usque ad 12 mm diametro, cu- 
pulata demum aperientia. Hymenium 50-55 pm crassum, subhymenium 
40-45 pm crassum, excipulum 260-310 pm crassum. Sporae non visae. 

Cortex superior K non reagens (acidum usnicum continens), medulla 
alba K non reagens, C non reagens, KC pallide rosea (acidum evernicum 
continens). 

Typus: CAPE — 3318 (Cape Town): Platteklip, Vlottenberg (-AD), on granite rock, 
S. Garside 5035(a) (BOL, holotype; US, isotype). 

Other Specimen Seen 

CAPE — 3218 (Clanwilliam): 25 km south of Clanwilliam, Olifants River Valley 
(-BD), on TMS, alt. 400-600 m, Brusse 772 8-1-6. 

Xanthoparmelia dysprosa is the seventh species in the X. hypoleia com- 
plex, which includes Parmelia ( Xanthoparmelia ) burmeisteri Elix, X. hypo- 
protocetrarica (Kurok. et Elix) Hale, X. notata (Kurok.) Hale, P. (Xantho- 
parmelia) pseudohypoleia Elix, X. cedrus-montana , described above, and X. 
karoo Knox et Brusse and X. leucostigma Brusse, described later. Within 
the group there is considerable uniformity of morphology between species, 
although some small variations in lobe size do occur. 

The chief common characteristic is the effigurate maculate appearance of 
the upper surface, and differentiation between species is based primarily on 
the different chemical constituents found in the medulla. In X. dysprosa is 
found the relatively uncommon (in Xanthoparmelia) orcinol p-depside, ever- 
nic acid, while P. burmeisteri contains the G-orcinol p-depsides barbatic and 
4-0-demethyl barbatic acids. Four other species, X. hypoleia, X. hypoproto- 
cetrarica, X. karoo and P. pseudohypoleia contain related G-orcinol depsi- 
dones, and X. notata contains the “mixed” depsidone, notatic acid. A fur- 
ther orcinol p-depside, gyrophoric acid, has been found in X. leucostigma, 
described subsequently. 


New Xanthoparmeliae from southern and central Africa 


149 




Fig. 3. 

Xanthoparmelia cylindriloba Knox (holotype). 


Fig. 4. 

Xanthoparmelia dysprosa Brusse et Knox (holotype). Scales in mm. 




150 


Journal of South African Botany 


Xanthoparmelia exornata (Zahlbr.) Brusse et Knox, comb. nov. Fig. 5. 
Basionym: Parmelia conturbata var. exornata Zahlbruckner (1932) Ann. 
Crypt. Exot. 5: 251-252. 

Typus: CAPE — Namaqualand: Steinkopf, Leg. Pastor G. Mayer, P. A. v. d. Bijl 950 
(W(1934) 315, holotype!; STEU, isotype). 

Thallus foliose, loosely to tightly adnate on rock, up to 100 mm across. 
Lobes plane to more usually strongly convex, sublinear to elongate, discrete 
to imbricate, 1-3 mm broad, leathery. Upper surface yellowish green, dark- 
ening towards the centre, strongly maculate to almost pseudocyphellate in 
regions where the epicortex is highly pored. Lower surface pale brown, 
moderately rhizinate. Upper cortex very uneven, 10-250 yarn thick; algal 
layer discontinuous, up to 100 /am thick; medulla 100-600 /am thick; lower 
cortex 20-40 /am thick. Apothecia sparse to moderately abundant, substipi- 
tate, up to 7 mm in diameter, shallowly cupped. Hymenium 40-60 /am 
thick, subhymenium 5-15 /am thick, exciple 30-120 /am thick. Ascospores 8, 
7,0-11,0 x 5, 0-6, 5 jam . Pycnidiospores 5-8 /am long. 

Upper cortex K“ (usnic acid), medulla K + yellow turning red (salazinic 
acid usually together with the chalybaeizans unknown). 

This very distinctive species is widespread in the Karoo, but shows no 
close relationships with any other Xanthoparmeliae. All the other strongly 
maculate species are in the X. hypoleia complex, and have narrower, more 
linear lobes, with a black undersurface and sparse rhizines. X. exornata rep- 
resents the extreme state of the effigurate maculate condition found in the 
X. hypoleia complex and in Omphalodium hottentottum (Ach.) Flot. In X. 
exornata the epicortical pores are closely aggregated, and may weather to 
produce discontinuities which resemble pseudocyphellae, but as pointed out 
by Hale (1981), investigation of the development of these structures will re- 
veal their true nature. 

Xanthoparmelia fucina Knox, sp. nov. Fig. 6. 

Thallus sordide flavescenti-virescens, saxicola adnatus vel laxe adnatus, 
centrum versus dense isidiatus. Lobi subirregulares, aliquantum dissecti, 
0, 7-2,0 mm lati. Pagina infera pallide castanea, interdum apices versus 
atrans. Rhizinae moderate abundantes cum pagina infera concolorae. Isidia 
plerumque simplicia, 0, 1-0,3 x 0,2-0, 7 mm. Cortex superior 18-20 /urn 
crassus, stratum gonidiale 30-35 /am crassum, medulla alba 75-90 /am cras- 
sa, cortex inferior 10-12 pm crassus. Apothecia non visa. 

Cortex superior K non reagens (acidum usnicum continens), medulla K, 
P aurantiaca (acidum protocetraricum et maculae irregulares rhodophysci- 
nae continens). 


New Xanthoparmeliae from southern and central Africa 15 1 



Xanthoparmelia exornata (Zahlbr.) Brusse et Knox. A. Typical growth form; B. The 
maculate upper surface. Scales in mm. 



152 


Journal of South African Botany 


Typus: ZIMBABWE: World’s View, Matopos, S Rhodesia, common on vertical to 
horizontal granite faces, usually east aspect, 9 vii 1946, E.A. Schelpe 1591 (BOL, 
holotype; US, isotype). 

Xanthoparmelia fucina is similar in form to X. subramigera (Gyel.) Hale, 
but differs in having slightly narrower lobes, and in containing protocetraric 
acid, rather than fumarprotocetraric and succinprotocetraric acids. Also 
similar morphologically is X. weberi (Hale) Hale, which has narrower lobes, 
smaller isidia and contains hypoprotocetraric acid. No sexual (apothecial) 
counterpart for X. fucina has been found. 

Xanthoparmelia hedbergii Knox, sp. nov. Fig. 7. 

Thallus sordide flavescenti-virescens, saxicola, arete adnatus, transverse 
rimosus vel centrum versus aliquantum areolatus. Lobi sublineares, 0, 5-1,0 
mm lati, pagina infera pallida vel badia, rhizinis paene apicibus; rhizinae 
cum pagina infera concolorae. Isidia cylindrica vel parum inflata, 0,05-0,1 x 
0, 1-0,3 mm abundantia sed non densa. Cortex superior 15-20 pm crassus, 
stratum gonidiale 30-35 gm crassum, medulla alba 100-130 pm crassa, cor- 
tex inferior 10-12 pm crassus. Apothecia non visa. 

Cortex superior K non reagens (acidum usnicum continens), medulla 
alba K non reagens, C non reagens vel pallide rosea, KC pallide rosea, P 
non reagens (acidum hypoprotocetraricum continens). 

Typus: KENYA: Rift Valley, S of Lake Naivasha, Olkaria area, near stream exit on 
recent lava flow, c2000 m, 5 iv 1976, 0. Hedberg s.n.. Flora of Kenya 6204d (UPS, 
holotype; US, isotype). 

Xanthoparmelia hedbergii is morphologically close to X. neocongensis 
(Hale) Hale and to X. weberi , and all three species contain hypoprotocetra- 
ric acid. Lobe size in X. neocongensis and X. hedbergii are in the same 
range, but the former species is distinguished by having a black lower sur- 
face. Lobes in X. weberi are broader than those in X. hedbergii. Although 
lobe size is influenced by environment, no tendency to a reduced lobe width 
could be found in South African populations of X. weberi, even in adverse 
conditions. On this basis, X. hedbergii is recognised as distinct from X. we- 
beri, and is assumed to have a more northerly distribution. 

Xanthoparmelia ianthina Brusse, sp. nov. Fig. 8. 

Thallus foliosus, flavovirens vel olivaceus, saxicola moderate adnatus, 
usque ad 60 mm in diametro. Lobi sublineares, 0, 5-2,0 mm lati; cortice 
superiore fatiscenti medulla ianthina revelata. Isidiis sorediisque carens. 
Pagina infera pallide spadicea, rhizinis sparsis vel moderate abundantibus. 
Cortex superior 12-20 pm crassus, stratum gonidiale 20-50 pm crassum, 
medulla 20-140 pm crassa, cortex inferior 5-15 pm crassus. Apothecia rara. 


New Xanthoparmeliae from southern and central Africa 


153 




Fig. 7. 


Xanthoparmelia hedbergii Knox (holotype). Scales in mm. 


Fig. 6. 

Xanthoparmelia fucipa Knox (holotype). 


it* 


154 


Journal of South African Botany 


substipitata, usque ad 2,5 mm in diametro cupulis vadosis. Hymenium 40-60 
pm crassum, subhymenium 20 /u,m crassum, excipulum 20-30 /u,m crassum. 
Sporae 8, 7,0-11,0 x 4,0-6, 5 /am; pycnidiosporae 5, 0-7,0 /am longae. 

Cortex superior K non reagens (usnic acid continens), medulla ianthina, 
K intense purpurea (pigmenta anthraquinonica et materia ignota continens). 

Typus: CAPE — 3119 (Calvinia): 42 km NE of Vanrhynsdorp, Vanrhyn’s Pass (-AC), 
on TMS, alt 460-610 m, Brusse 768-10-1-15. 

This species closely resembles X. endomiltodes (Nyl) Hale, but differs in 
lacking salazinic acid and in containing two unknown substances, designated 
Th-1 and Th-2. These two substances are suspected (3-orcinol m-depsides, 
exhibiting the following characteristics. Th-1 is K + yellow, rapidly becoming 
bright orange-red. On thin-layer chromatography (TLC) plates developed in 
10 % sulphuric acid it is olive to brown, whereas Th-2 is olive to greyish. 
Th-1 is yellow-brown to brown in long-wave (366 nm) uv light, and Th-2 is 
brown. Table 1 gives TLC data for these two substances in comparison with 
the standards atranorin and norstictic acid, expressed in the format em- 
ployed by Culberson (1972). 

Microhydrolysis of Th-1 according to the method described by Culber- 
son (1972), and using for comparison baeomycesic acid obtained from Si- 
phula torulosa (Thunb. ex Ach.) Nyl. indicated that the A-ring in both cases 
was (after oxidation by sulphuric acid during the hydrolysis procedure) 
4-0-methyl-3-carboxyorsellinic acid. No second subunit could be detected 
from this procedure, and it was assumed to have remained at the origin. 

Xanthoparmelia karoo Knox et Brusse, sp. nov. Fig. 9. 

Thallus foliosus, saxicola, moderate vel arete adnatus, usque ad 60 mm 
in diametro. Lobi 0,8-1, 5 mm lati, convexi, lineares vel sublineares, unifor- 


Table 1. 

TLC data for the two suspected m-depsides Th-1 and Th-2, using norstictic acid 
and atranorin as standards. 


Lichen 

substance 

Rf class 
Solvents 

R f x 100 a 
Solvents 


A 

B 

C 

A 

B 

C 

Th-1 

4 

1-2 

3 

40/40,75 (56) 

6/39,73 (56) 

18/34,77 (55) 

Th-2 

3 

1 

1-2 

21/35,72 (54) 

0/35,65 (49) 

6/32,77 (54) 


The Rf X 100 value is given, followed by a virgule, and then the Rf X 100 values 
for norstictic acid and atranorin, with the mean of these two in parentheses. 


New Xanthoparmeliae from southern and central Africa 


155 


miter appressi, dilute vel manifeste albomaculati. Pagina infera nigra, grosse 
rhizinata. Isidiis sorediisque carens. Cortex superior 20-100 /am crassus, 
stratum gonidiale 40-80 gm crassum, medulla alba 50-500 /u,m crassa, cor- 
tex inferior 30-60 /am crassus. Apothecia sparsa vel moderate abunda, sub- 
stipitata, usque ad 5 mm diametro, cupulis vadosis. Hymenium 40-60 /am 
crassum, subhymenium 5-10 /am crassum, excipulum 30-60 /am crassum. 
Ascosporae 8, 8,0-13,0 x 4, 0-6,0 /am. Pycnidia 100-200 /am diametro, 
pycnidiosporae 5, 0-8,0 /am longae. 

Cortex superior K non reagens (acidum usnicum continens), medulla K 
non reagens, C pallide rosea, KC rosea (acidum hypoprotocetraricum conti- 
nens, aliquando acido 4-0-methylhypoprotocetrarico). 

Typus: CAPE — 3219 (Wuppertal): 32 km NE of Clanwilliam, Klipfonteinrand 
(-AA), on TMS, alt. 300-450 m, Brusse 768-10-3-7 (J, holotype). 

Morphologically, X. karoo resembles X. exornata , but is less distinctly 
maculate. The medullary chemistry is identical to that of X. hypoprotocetra- 
rica, but the latter species is less adnate, has broader lobes and usually ex- 
hibits a denser, more imbricate or even pulvinate growth form. In lobe 
form, X. karoo may be considered as part of the X. hypoleia complex, but is 
distinguished from the other members of this group by its closely adnate 
mode of growth. 

Xanthoparmelia leucostigma Brusse, sp. nov. Fig. 10. 

Thallus sordide flavovirescens, saxicola, moderate vel arete adnatus. 
Lobi lineares, plani vel convexi, uniformiter appressi, contigui, 0,5-3,0 mm 
lati, albomaculati. Pagina infera eborina vel pallide fusca, moderate rhizina- 
ta, rhizinis grossis. Isidiis sorediisque carens. Cortex superior 15-70 /am 
crassus, stratum gonidiale irregulare, 40-80 /am crassum, medulla 60-200 
/am crassa, cortex inferior 15-30 /am crassus. Apothecia sparsa vel moderate 
numerosa, substipitate, usque ad 3 mm diametro. Hymenium 40-60 /am 
crassum, subhymenium 10-15 /am crassum, excipulum 20-40 /am crassum. 
Ascosporae 8, 6, 0-9, 5 x 4, 5-5, 5 /am. Pycnidia globosa, 100-200 /am di- 
ametro, pycnidiosporae 5-8 /am longae. 

Cortex superior K non reagens (acidum usnicum continens), medulla 
alba K non reagens, C dilute rosea (acidum gyrophoricum continens). 

Typus: CAPE— 3322 (Oudtshoorn): 18 km N of De Rust, Meiringspoort (-BC), F. 
Brusse 772 14-1-17 (PRE, holotype). 

This species is at present known only from the type collection, but is im- 
portant in that it is a further member of the X. hypoleia group containing 
another orcinol p-depside, gyrophoric acid. Chemical relationships within 
this group will be discussed in a subsequent publication. 


156 


Journal of South African Botany 



Fig. 8. 

Xanthoparmelia ianthina Brusse (holotype). 



Fig. 9. 

Xanthoparmelia karoo Knox et Brusse (holotype). Scales in mm. 






New Xanthoparmeliae from southern and central Africa 157 

Xanthoparmelia walteri Knox, sp. nov. Fig. 11. 

Thallus foliosus, flavovirens, prostratus vel subascendens, usque ad 30 
mm diametro, saxa parva habitans. Lobi ligulati, maculatuli, 0,5-1, 5 mm 
lati, paginis inferibus nigris, rugosis, nudis vel sparse rhizinatis. Isidiis sore- 
diisque carens. Cortex superior 18-20 gm crassus, stratum gonidiale 30-35 
/xm crassum, medulla 90-110 gm crassa, cortex inferior 12-15 gm crassus. 
Apothecia non visa. 

Cortex superior K non reagens (acidum usnicum continens), medulla 
alba K flavescens turn sanguinea (acidum salazinicum et acidum norsticticum 
minutissimum continens). 

Typus: S.W. A. /NAMIBIA — 2214 (Swakopmund): Namibwiiste 10 bis 20 km 
nordlich von Swakopmund (-DAj, Kieswuste auf dunklen Doleritsteinchen, Mitte 
Oktober 1977 + December 1977, L. Moisei fur H. and E. Walter No. 5168. 

Salazinic acid is certainly the most common medullary constituent in 
Xanthoparmelia , there being more than 50 valid species containing this sub- 
stance. Xanthoparmelia walteri , lacking isidia and soredia, is clearly related 
to X. constrictans (Nyl.) Hale, but differs in its semi-erect growth form and 
in its more regular, straplike lobes. 



Fig. 10. 

Xanthoparmelia leucostigma Brusse (holotype). Stales in mm. 


158 


Journal of South African Botany 



Fig. 11. 

Xanthoparmelia walteri Kriox (holotype). Scale in mm. 


Acknowledgements 

The authors are grateful to the University of the Witwatersrand for 
financial support during the completion of the degree of M.Sc. by one of 
them (F.A.B.), and to the Council for Scientific and Industrial Research and 
the above University for financial support during a period of overseas study 
leave for the other (M.D.E.K.). 

Thanks are also due to Dr. Mason E. Hale, Jr., of the Smithsonian Insti- 
tution, Washington, D.C. for guidance during this work. 

References 

Beltman, H. A., 1978. Vegetative Strukturen der Parmeliaceae und ihre Entwick- 
lung. Biblthca. lichenol., Lehre 11 : 1-193. 

Culberson, C. F., 1972. Improved conditions and new data for the identification of 
lichen products by a standardised thin-layer chromatographic method. J. 
Chromat. 72 : 113-125. 

Elix, J. A., 1981. New species of Parmelia subgen. Xanthoparmelia (Lichens) from 
Australia and New Zealand. Aust. J. Bot. 29 : 349-376. 


New Xanthoparmeliae from southern and central Africa 159 

Hale, M. E., Jr., 1974. Bulbothrix, Parmelina, Relicina and Xanthoparmelia, four 
new genera in the Parmeliaceae (Lichenes). Phytologia 28: 479-490. 

Hale, M. E., Jr., 1981. Pseudocyphellae and pored epicortex in the Parmeliaceae: 
their delimitation and evolutionary significance. Lichenologist 13: 1-10. 

Hale, M. E., Jr. and Kurokawa, S., 1964. Studies on Parmelia subgenus Parmelia. 
Contr. U.S. natn. Herb. 36: 121-191. 

Hawksworth, D. L., James, P. W. and Coppins, B. J., 1980. Checklist of British li- 
chen-forming, lichenicolous and allied fungi. Lichenologist 12: 1-115. 

Knox, M. D. E., 1981. Chemical and morphological variation in the Xanthoparmelia 
hypoleia (Nyl.) Hale group (Lichenes). Annual Congress of the South Afri- 
can Association of Botanists, Port Elizabeth, South Africa, January, 1981. 
p.24. 

Krog, H. and Swlnscow, T. D. V., 1979. Parmelia subgenus Hypotrachyna in East 
Africa. Norw. J. Bot. 26: 11-43. 

Lynge, B., 1937. Lichenes Africani novi ab E. A. Vainio recogniti. Rev. bryol. li- 
chen. 10: 78-91. 

Nylander, W., 1860. Synopsis Methodica Lichenum. Vol. 1. Paris: L. Martinet. 

Poelt, J. and V£zda, A. 1981. Bestimmungschliissel europaischer Flechten. Ergan- 
zungsheft II. Vaduz: J. Cramer. 



J1 S. Afr. Bot. 49 (2): 161-174 (1983) 


AN ANNOTATED SYSTEMATIC CHECKLIST OF THE 
ANGIOSPERMAE OF THE CAPE RECEIFE NATURE RESERVE, PORT 
ELIZABETH 

Maria C. Olivier 

(Department of Botany, University of Port Elizabeth , P.O. Box 1600, Port 
Elizabeth 6000, R.S.A.) 

Abstract 

The vegetation of the Cape Receife Nature Reserve, Port Elizabeth is discussed 
briefly. The 173 species recorded are listed in a systematic checklist. Of the 60 famil- 
ies recorded, the Asteraceae (Compositae) is the one best represented by far. 

UlTTREKSEL 

’N GEANNOTEERDE SISTEMATIESE KONTROLELYS VAN DIE ANGIO- 
SPERMAE VAN DIE KAAP RECEIFE-NATUURRESERVAAT, PORT ELI- 
ZABETH 

Die plantegroei van die Kaap Receife-Natuurreservaat, Port Elizabeth, word 
kortliks bespreek. Die 173 spesies wat aangeteken is, word in ’n sistematiese kontro- 
lelys weergegee. Van die 60 families wat aangeteken is, is die Asteraceae (Composi- 
tae) op verre na die beste verteenwoordig. 

Introduction 

The Cape Receife Nature Reserve (S 34° 01 E 25° 42 ') is a small re- 
serve, 366 ha in extent and situated 9,6 kilometres from the centre of the 
city of Port Elizabeth. It has approximately six kilometres of coastline and 
on the landward side it is bounded by the campus of the University of Port 
Elizabeth and the South African Defence Force rifle range. 

The terrain is undulating, the highest point in the area, Receife Hillock, 
being 45 metres above sea level. Running in an east-west direction and sep- 
arating the peninsula from the mainland is a broad belt of dune sand. In the 
eastern half of this belt a sewage treatment plant, with settling ponds, is well 
established. It is also in this half that an area of shifting sand has been stabi- 
lized successfully, a grass and shrub mixture having been used for this pur- 
pose. A few natural freshwater ponds also occur in the area. 

Three roads traverse the area giving access to the lighthouse, the sewage 
works and the car park at the western side of the reserve. 

The area was proclaimed a nature reserve in 1973 and is administered by 


Accepted for publication 20th September, 1982. 


161 


162 


Journal of South African Botany 


the Parks and Recreation Department of the City Council of Port Elizabeth. 
Entry is by permit only. 

The survey was carried out because a reasonable knowledge of species 
occurring in a protected area is essential and very little floristic data were 
available. 

Methods 

During 1981 the area was visited at regular intervals, but also at odd 
times prior to this. Specimens were collected in duplicate. The first set is 
housed in the herbarium of the University of Port Elizabeth and the second 
set was sent to the Botanical Research Unit, Grahamstown, where it was 
used by the author and the staff of the unit for identification purposes. 

Plant communities were assessed visually. Flowering periods were re- 
corded as well as presence or absence of species in the different communi- 
ties. Abundance was estimated on a five-point scale: rare, uncommon, fairly 
common, common and very common. 

Results and Discussion 

Vegetation 

Acocks (1975) considers the area to be a coastal tropical forest type. He 
classifies it as Alexandria Forest (Veld Type 2) but does not discuss this re- 
lationship in the text and only shows it on the map. 

During the present survey six communities were distinguished (Fig. 1). 

1. Pioneer communities on unstabilized dunes, more or less concentrated 
along the eastern coastline. 

The most important pioneers are Arctotheca populifolia and Scaevola 
thunbergii. 

2. Rocky shore communities which occur on consolidated areas between the 
rocky shore and dune scrub, along the southern and south-western coast- 
line. 

The most important species occurring here are: Chenolea diffusa, Cotula 
species, Gazania rigens (both varieties), Limonium linifolium, Sporo- 
bolus virginicus and Tetragonia decumbens. 

3. Dune scrub. This community occupies the largest area by far in the re- 
serve. It is virtually impenetrable and collecting was limited to roadsides 
and an area that was destroyed by fire during February 1978. 

The dominant species are: Cassine species, Olea exasperata, Pterocelas- 
trus tricuspidatus, Rhus species and Sideroxylon inerme. 

4. Fynbos. Although it is not impossible that several fynbos patches are to 
be found within the dune scrub, only one of limited extent was easily ac- 
cessible and investigated. 


2 Rocky shore- 


Checklist of Angiospermae of Cape Receife Nature Reserve 


163 



164 


Journal of South African Botany 


The commonest species occurring here are: Agathosma apiculata, Mural- 
da squarrosa, Myrica quercifolia, Passerina vulgaris and Sutera micro- 
phylla. 

5. Freshwater pond community. A few natural freshwater ponds are to be 
found in the area. 

Typha latifolia subsp. capensis is the dominant species. 

6. Dry limestone pan community situated near the settling ponds. 

The most important species are: Chondropetalum microcarpum and My- 
rica cordifolia. 

Nothing comparable to dune forest occurs in the reserve. 

Dense stands of Acacia cylops, especially in the northern and eastern 
parts of the reserve, are a serious threat to the indigenous vegetation. 

Flora 

During the survey, 60 families, 127 genera and 173 species of angio- 
sperms were listed. An analysis of the families indicates that 11 (18,33 %) 


Table 1. 

Synopsis of the families whose species comprise more than 2 % of the total number 
listed in order of numerical importance together with the number of genera. 


Family 

No. of 
species 

No. of 
species 
expressed as 
a % of total 

No. of 
genera 

No. of 
genera 
expressed as 
a % of total 

Asteraceae (Compositae) . . 

28 

16,18 

15 

11,81 

Poaceae (Gramineae) 

11 

6,36 

10 

7,87 

Liliaceae 

10 

5,78 

8 

6,30 

Cyperaceae 

9 

5,20 

4 

3,15 

Fabaceae (Leguminosae) . . 

9 

5,20 

6 

4,72 

Celastraceae 

8 

4,62 

4 

3,15 

Mesembryanthemaceae .... 

6 

3,47 

6 

4,72 

Chenopodiaceae 

5 

2,89 

4 

3,15 

Crassulaceae 

4 

2,31 

2 

1,57 

Iridaceae 

4 

2,31 

2 

1,57 

Santalaceae 

4 

2,31 

4 

3,15 


Checklist of Angiospermae of Cape Receife Nature Reserve 


165 


are monocotyledons, and 49 (81,67 %) dicotyledons. The first is represented 
by 43 (24,86 %) species and the latter by 130 (75,14 %) species. There are 
11 (18,33 %) families whose species contribute more than 2 % of the total 
number of species and these are listed in order of numerical importance in 
Table 1. Twenty nine (48,33 %) families are represented by one species, 14 
(23,33 %) by two species and six (10,00 %) by three species. 

Phenology 


Flowering periods for 164 species were recorded. An analysis of these in- 
dicates a remarkable ascending fluctuation with two major peaks in Septem- 
ber and November (Fig. 2). 



Months 


Histogram showing the relationship between the species flowering and the months of 

the year. 


Systematic List 

The systematic list is arranged according to Dyer (1975, 1976), with the 
exception of Viscum which is classified according to Wiens and Tolken 
(1979). The numbers refer to the author’s collecting number; fl. = flowering 
period; 1-12 = months of the year. 


166 


Journal of South African Botany 


MONOCOTYLEDONAE 

TYPHACEAE 

Typha latifolia L. subsp. capensis Rohrb. 1944, fl. 11, freshwater ponds, common. 
JUNCAGINACEAE 

Triglochin striata Ruiz & Pav. 2400, fl. 11, seepage of freshwater ponds, fairly com- 
mon. 


POACEAE (Gramineae) 

Cymbopogon marginatus (Steud.) Stapf ex Burtt Davy 2377, fl. 9-11, dune scrub, 
burnt area, uncommon. 

Stenotaphrum secundatum (Walt.) Kuntze 3030, fl. 12, rocky shore community, 
fairly common. 

Ehrharta calycina J. E. Sm. 2983, fl. 9-11, dune scrub, burnt area, uncommon. 

E. villosa Schult. f. 1887, fl. 9-11, pioneer dune communities, fairly common. 

Pentaschistis heptomera (Nees) Stapf 2403, fl. 9-11, pioneer dune communities, un- 
common. 

Lagurus ovatus L. 1932, fl. 9-11, dune scrub, disturbed areas, fairly common. 

Sporobolus virginicus (L.) Kunth 2742, fl. 12-1, rocky shore communities, fairly 
common. 

Cynodon dactylon (L.) Pers. 3033, fl. 12, rocky shore community, uncommon. 

Lolium multiflorum Lam. 3006, fl. 10-11, dune scrub, disturbed areas, common. 

Agropyron distichum (Thunb.) Beauv. 2741, fl. 11-1, pioneer dune communities, 
common. 

Setaria sp. 2398, fl. 11, dune scrub, disturbed areas, uncommon. 


CYPERACEAE 

Juncellus laevigatus C.B.C1. 3013, fl. 12, damp area beyond freshwater pond, local- 

ly fairly common. 

Ficinia cf. F. aphylla Nees 2782, fl. 2, dry limestone pans, fairly common. 

F. bulbosa Nees 2380, fl. 11, dune scrub, burnt area, common. 

F. dunensis Levyns 2925, fl. 6, fynbos, uncommon. 

F. cf. F. ramosissima Kunth 1872, fl. 9, dry limestone pans, fairly common. 

F. truncata Schrad. 2273, fl. 8, dune scrub, burnt area, common. 

Fuirena cf. F. hirsuta (Berg.) P. L. Forbes 3014, fl. 12, damp area beyond fresh- 
water pond, locally fairly common. 

Scirpus nodosus Rottb. 1937, fl. 11, pioneer dune communities near entrance to re- 
serve, fairly common. 

S. thunbergianus (Nees) Levyns 3015, fl. 12, damp area beyond freshwater pond, 
locally fairly common. 


ARACEAE 

Zantedeschia aethiopica (L.) Spreng. 2763, fl. 6-2, freshwater ponds, fairly com- 
mon. 


Checklist of Angiospermae of Cape Receife Nature Reserve 


167 


RESTIONACEAE 

Chondropetalum microcarpum Pillans 1869, 2783, fl. 2, 9, dry limestone pans, com- 
mon. 


JUNCACEAE 

Juncus kraussii Elochst. 3037, fl. 11—12, pioneer dune communities near entrance 
gate, common. 


LILIACEAE 

Androcymbium longipes Bak. 1849, fl. 9, dune scrub, fairly common. 

Bulbine caulescens L. 1946, fl. 11, dune scrub, fairly common. 

Trachyandra ovatum Kies 2986, fl. 8-9, dune scrub, fairly common, wide 
strapshaped leaves. 

T. revolutum L. 2984, fl. 8-9, dune scrub, fairly common, narrow strapshaped 
leaves. 

Aloe africana Mill. 2950, fl. 5-6, scattered in dune scrub. 

Gasteria croucheri Bak. 2366, fl. 11, scattered in dune scrub. 

Albuca sp. 2382, fl. 10-11, dune scrub, uncommon. 

Massonia echinata L.f. 2959, 2985, fl. 6-7, dune scrub, burnt area, fairly common. 
Asparagus asparagoides (L.) Wight 2930, fl. 6, dune scrub, fairly common. 

A. racemosus Willd. 3032, fl. not recorded, dune scrub, fairly common. 

AMARYLLIDACEAE 

Boophane disticha Herb. 2780, fl. 1-2, dune scrub, burnt area, fairly common. 


IRIDACEAE 

Chasmanthe aethiopica (L.) N.E.Br. 2828, fl. 5-6, dune scrub, common. 

Gladiolus floribundus Jacq. subsp. floribundus 3007, fl. 10, dune scrub, fairly com- 
mon. 

G. floribundus Jacq. subsp. milled (Ker) Oberm. 2381, fl. 11, very abundant in 
burnt area. 

G. guernzii Klatt 2390, fl. 11, pioneer dune community, uncommon. 


ORCHIDACEAE 

Holothrix exilis Lindl. 2373, fl. 11, fynbos, south-facing area, rare. 

Bonatea speciosa (L.f.) Willd. 2368, fl. 11, dune scrub, uncommon. 

Satyrium princeps Bol. 2987, fl. 9, stabilised dunes near car park, uncommon. 


DICOTYLEDONAE 


MYRICACEAE 

Myrica cordifolia L. 2785, fl. 9, dry limestone pan, fairly common. 
M. quercifolia L. 2497, fl. 8-9, fynbos, common. 


168 


Journal of South African Botany 


VISCACEAE 

Viscum obscurum Thunb. 2948, fl. 6, dune scrub, uncommon. 


SANTALACEAE 

Rhoiacarpos capensis (Harv.) A. DC. 1868, fl. 1-12, dune scrub, common. 

Colpoon compression Berg. 2797, fl. 1-12, dune scrub, fairly common. 

Thesidium exocarpaeoides Sond. 2786, 2973, fl. 2, 8, fynbos, fairly common. 

Thesium sp. 2764, fl. 2, rocky shore community, fairly common. 

CHENOPODIACEAE 

Chenopodium cf. C. album L. 2799, fl. 3, rocky shore community, uncommon. 

C. murale L. 2829, fl. 5, disturbed area along road, common. 

Exomis microphylla (Thunb.) Aell. var. axyrioides (Fenzl) Aell. 1866, fl. 9, dune 
scrub, uncommon. 

Chenolea diffusa Thunb. 3031, fl. not recorded, rocky shore community, common. 
Sarcocornia perennis (Miller) A. J. Scott 3029, fl. 12, rocky shore community, un- 
common. 


AIZOACEAE 

Aizoon rigidum L.f. 2927, fl. 6, fynbos, uncommon. 

Tetragonia decumbens Mill. 2923, fl. 6-9, rocky shore community, common. 
T. fruticosa L. 2992, fl. 9, pioneer dune communities, common. 


MESEMB RY ANTHEM ACE AE 

Carpobrotus deliciosus (L. Bol.) L. Bol. 2963, fl. 8-9, dune scrub, fairly common. 
Delosperma littorale (Kensit) L. Bob 2964, fl. 8, fynbos, fairly common. 

Disphyma crassifolia (L.) L. Bob 3028, fl. 8-2, rocky shore community, common. 
Drosanthemum intermedium L. Bob 2965, fl. 8, fynbos, fairly common. 
Mesembryanthemum aitonis Jacq. 2831, fl. 5, disturbed area in dune scrub, fairly 
common. 

Nycteranthus plenifolius (N.E. Br.) Schwant. 2363, fl. 11, dune scrub, burnt area, 
common. 


CARYOPHYLLACEAE 

Silene bellidioides Sond. 2979, fl. 9, rocky shore community, common. 

S. primulaeflora Eckl. & Zeyh. 2370, fl. 1-12, pioneer dune communities, com- 
mon. 

LAURACEAE 

Cassytha cf. C. filiformis L. 2778, fl. 2, dune scrub, fairly common. 
BRASSICACEAE (Cruciferae) 

Heliophila linearis (Thunb.) DC. var. linearis 1890, fl. 10, dune scrub, uncommon. 


Checklist of Angiospermae of Cape Receife Nature Reserve 


169 


CRASSULACEAE 

Cotyledon orbiculata L. 2803, fl. 11-3, dune scrub, fairly common. 

Crassula cotyledonis Thunb. 2364, fl. 11, dune scrub, common. 

C. expansa Dryand. subsp. filicaulis (Haw.) Toelk. 2387, fl. 10-11, dune scrub, 
common. 

C. sp. 2388, fl. 11, dune scrub, uncommon. 


FABACEAE (Leguminosae) 

Aspalathus lactea Thunb. subsp. adelphea (Eckl. & Zeyh.) Dahlg. 3018, fl. 11-2, 
fynbos, carpark turnoff, rare. 

Crotalaria capensis Jacq. 1861, fl. 9, dune scrub on Receife Hillock, rare. 

Indigo f era heterophylla Thunb. 2924, fl. 6, fynbos, fairly common. 

I. incana Thunb. var. angustistipulata Bak. f. 2970, 2989, fl. 8, fynbos and dune 
scrub, burnt area, fairly common. 

I. striata L. 815, fl. 9, fynbos, uncommon. 

I. sulcata DC. 2939, fl. 6, fynbos, common. 

Psoralea bracteata L. 1935, fl. 11, fynbos, common. 

Lessertia stenoloba E. Mey. 1938, 2375, fl. 9, fynbos and dune scrub, burnt area, 
fairly common. 

Rhynchosia caribaea (Jacq.) DC. var. picta (E. Mey.) Bak. f. 2275, fl. 8, dune 
scrub, burnt area, uncommon. 


GERANIACEAE 

Pelargonium capitatum (L.) Ait. 2996, fl. 9-11, rocky shore community, common. 

P. urbanum (Eckl. & Zeyh.) Harv. 2378, fl. 5-11, dune scrub, burnt area, com- 
mon. 


OXALIDACEAE 

Oxalis punctata L.f. 2818, fl. 4, rocky shore community, uncommon. 


LINACEAE 

Linum africanum L. 3002, fl. 10-11, fynbos, next to road at carpark turnoff, local- 
ised. 


ZYGOPHYLLACEAE 

Zygophyllum uitenhagense Sond. 2816, fl. 1-12, disturbed areas, common. 


RUTACEAE 

Agathosma apiculata G. Meyer 2968, fl. 8-9, fynbos, common. 
Coleonema pulchrum Hook. 2929, fl. 5-6, fynbos, fairly common. 


170 


Journal of South African Botany 


POLYGALACEAE 

Polygala ericaefolia DC. 3008, fl. 10, fynbos, uncommon. 

Muraltia squarrosa (L.f.) DC. 2958, fl. 4-11, fynbos, common. 
Nylandtia spinosa (L.) Dumort. 2931, fl. 6, dune scrub, uncommon. 


ANACARDIACEAE 

Rhus crenata Thunb. 2810, fl. 4, dune scrub, fairly common. 
R. glauca Thunb. 1852, fl. 6, dune scrub, fairly common. 

R. schlechteri Diels 2389, fl. 9, dune scrub, common. 


CELASTRACEAE 

Maytenus procumbens (L.f.) Loes. 2812, fl. 2-6, dune scrub, fairly common. 
Putterlickia pyracantha (L.) Szyszyl. 3019, fl. 11, dune scrub, fairly common. 
Pterocelastrus tricuspidatus (Lam.) Sond. 2945, fl. 11, dune scrub, common. 

Cassine aethiopica Thunb. 3021, fl. 4-11, dune scrub, fairly common. 

C. eucleiformis (Eckl. & Zeyh.) Kuntze 2278, fl. 8, dune scrub, common. 

C. peragua L. 2951, fl. 6, dune scrub, fairly common. 

C. reticulata (Eckl. & Zeyh.) Codd 1873, fl. not recorded, dune scrub, uncommon. 
C. tetragona (L.f.) Loes. 2775, fl. 2, dune scrub, fairly common. 

RHAMNACEAE 

Phylica ericoides L. 2792, fl. 3, fynbos, fairly common. 

VITACEAE 

Rhoicissus tridentata (L.f.) Wild & Drumm. 2798, fl. 3, dune scrub, fairly common. 
MALVACEAE 

Abutilon sonneratianum (Cav.) Sweet 1876, fl. 9, dune scrub, uncommon. 

Lavatera arborea L. 1860, fl. 9, weed next to road at carpark turnoff, only a few 
plants. 

STERCULIACEAE 

Hermannia althaeoides Link. 2530, fl. 8, fynbos, uncommon. 

H. filifolia L.f. 2952, fl. 6, fynbos, uncommon. 

FLACOURTIACEAE 

Dovyalis rotundifolia (Thunb.) Harv. 3022, fl. not recorded, Receife Hillock, rare. 


THYMELAEACEAE 

Passerina rigida Wikstr. 1893, fl. 10, pioneer dune communities, fairly common. 
P. vulgaris (Meisn.) Thoday 2988, fl. 9, fynbos, fairly common. 


Checklist of Angiospermae of Cape Receife Nature Reserve 


171 


ONAGRACEAE 

Oenothera drummondii Hook. 2977, fl. 9, next to roadside near entrance, common. 


ARALIACEAE 

Cussonia thyrsiflora Thunb. 2770, fl. 2, dune scrub, fairly common. 


APIACEAE (Umbelliferae) 

Heteroptelis suffruticosa (Berg.) Leute 822, fl. 1-12, edge of dune scrub, near gate 
to lighthouse, uncommon. 


ERICACEAE 

Erica chloroloma Lindl. 2809, fl. 4-11, fynbos, fairly common. 

E. cf. E. glumaeflora Klotzsch ex Benth. 2936, fl. 6-8, fynbos, fairly common. 


MYRSINACEAE 

Rapanea gilliana (Sond.) Mez. 2940, fl. 6-8, dune scrub, fairly common. 


PRIMULACEAE 

Samolus porosus (L.f.) Thunb. 2765, fl. 2, swampy area in north-eastern part of re- 
serve, uncommon. 


PLUMBAGINACEAE 

Limonium linifolium (L.f.) Kuntze 3035, fl. 3-11, rocky shore community, fairly 
common. 


SAPOTACEAE 

Sideroxylon inerme L. 2772, fl. 11-12, dune scrub, common. 


EBENACEAE 

Euclea natalensis A. DC. 2949, fl. 6, dune scrub, fairly common. 

E. racemosa Murr. subsp. bernardii F. White 1853, fl. not recorded, dune scrub, 
fairly common. 


OLEACEAE 

Olea exasperata Jacq. 2941, fl. 6-11, dune scrub, common. 


SALVADORACEAE 

Azima tetracantha Lam. 2947, fl. not recorded, dune scrub, Receife Hillock, un 
common. 


172 


Journal of South African Botany 


GENTIANACEAE 

Chironia baccifera L. 3016, ft. 11-12, fynbos, common. 

C. decumbens Levyns 3036, fl. 12-3, dry limestone pan, uncommon. 

C. tetragona L.f. 2495, fl. 2-3, dry limestone pan, uncommon. 

APOCYNACEAE 

Carissa bispinosa (L.) Desf. ex Brenan 2982, fl. 9, dune scrub, fairly common. 


ASCLEPIADACEAE 

Astephanus marginatas Decne 2830, fl. 5-11, dune scrub, common. 

Secamone alpinii Schult. 1892, fl. 10-11, dune scrub, common. 

CONVOLVULACEAE 

Ipomoea cairica (L.) Sweet 2819, fl. 1-12, near freshwater ponds, common. 
LAMIACEAE (Labiatae) 

Leonotis dubia E. Mey. 2372, fl. 6-11, dune scrub on Receife Hillock, uncommon. 
Salvia africana-lutea L. 2993, fl. 9-10, dune scrub, uncommon. 

SOLANACEAE 

Lycium tetrandrum Thunb. 2796, fl. 3-10, rocky shore community, fairly common. 
Solanum quadrangulare Thunb. 2791, fl. 3-9, dune scrub, common. 

SCROPHULARIACEAE 

Sutera campanulata (Benth.) Kuntze 2794, fl. 3-8, dune scrub, burnt area, com- 
mon. 

S. microphylla (L.f.) Hiern 2820, fl. 4-11, fynbos, common. 

Zaluzianskya capensis Walp. 2953, fl. 6-11, dune scrub, burnt area, fairly common. 
SELAGINACEAE 

Hebenstreitia cordata L. 1947, fl. 11, near entrance to reserve, uncommon. 
PLANTAGINACEAE 

Plantago carnosa Lam. 2385, fl. 11, rocky shore community, common. 
RUBIACEAE 

Anthospermum aethiopicum L. 2971, fl. 8, fynbos, fairly common. 

Hydrophylax carnosa Sond. 1885, fl. 10, found on one dune only, west of light- 
house. 

CUCURBITACEAE 

Zehneria scabra (L.f.) Sond. 3012, fl. 10, dune scrub, uncommon. 

Kedrostis nana (Lam.) Cogn. var. nana 2771, fl. 9-11, dune scrub, fairly common. 


Checklist of Angiospermae of Cape Receife Nature Reserve 


173 


GOODENIACEAE 

Scaevola thunbergii Eckl. & Zeyh. 3023, fl. 11-2, pioneer dune communities, com- 
mon. 


ASTERACEAE (Compositae) 

Felicia echinata (Thunb.) Nees 2832, fl. 5-9, fynbos, uncommon. 

F. erigeroides DC. 2928, fl. 5-6, dune scrub, uncommon. 

Conyza pinnatifida Less. 2766, fl. 2, swampy area in north-eastern part of reserve, 
uncommon. 

C. scabrida DC. (=C. ivifolia (L.) Less.) 2767, fl. 2, swampy area in north-eastern 
part of reserve. 

Tarchonanthus camphoratus L. 2793, fl. 2-4, dune scrub, uncommon. 

Helichrysum cymosum (L.) D. Don 2801, fl. 2-3, swampy area in north-eastern 
part of reserve. 

H. sordescens DC. 1845, fl. 9, dune scrub, common. 

H. teretifolium (L.) D. Don 1844, fl. 9-10, dune scrub, common. 

Stoebe cinerea (L.) Thunb. 1870, fl. 9, near settling ponds, uncommon. 

Metalasia muricata (L.) D. Don 1305, fl. 5-6, widespread, fairly common. 

Cotula coronopifolia L. 2955, fl. 6-11, seepage area near freshwater ponds, uncom- 
mon. 

C.sp.(=Cenia pectinata (Eckl.) DC.) 2967, fl. 8-11, rocky shore communities and 
dune scrub in open spaces, common. 

C.sp.(=Cem'a sericea DC.) 2921, fl. 6-8, rocky shore communities and dune scrub 
in open spaces, common. 

Pentzia globifera (Thunb.) Hutch. 2402, fl. 11, disturbed area in dune scrub, com- 
mon. 

Senecio elegans L. var. diffusus Harv. 2975, fl. 9-10, pioneer dune community, fair- 
ly common. 

S. ilicifolius Thunb. 3009, fl. 10, disturbed area in dune scrub, common. 

5. inaequidens DC. 2833, fl. 5-10, disturbed area in dune scrub, common. 

5. lanceus Ait. 2768, fl. 2, swampy area in north-eastern part of reserve, uncom- 
mon. 

5. oederiaefolia DC. 2978, fl. 9-10, rocky shore community, common. 

Othonna amplexicaulis Thunb. 2956, fl. 4-9, dune scrub, common. 

Osteospermum fruticosum (L.) T. Norl. 2974, fl. 8-9, rocky shore community, fair- 
ly common. 

Chrysanthemoides monilifera (L.) T. Norl. 1302, fl. 1-12, widespread. 

Ursinia chrysanthemoides Harv. 2990, fl. 8-11, dune scrub, burnt area, common. 

Arctotheca calendula (L.) Levyns 2836, fl. 5, disturbed area along road to light- 
house, common. 

A. populifolia (Berg.) T. Norl. 1299, fl. 1-12, pioneer dune communities, common. 

Gazania linearis (Thunb.) Druce 3034, fl. 8-2, fynbos, fairly common. 

G. rigens (L.) Gaertn. var. leucolana (DC.) Roessl. 2485, fl. 1-12, rocky short 

community, common, leaves hairy on both surfaces. 

G. rigens (L.) Gaertn. var. uniflora (L.f.) Roessl. 2815, fl. 1-12, rocky shore com- 
munity, common, leaves hairy on lower surface. 


174 


Journal of South African Botany 


Acknowledgements 

I wish to thank the City Council of Port Elizabeth for permission to 
undertake this survey; the staff of the Albany Museum Herbarium, Botan- 
ical Research Unit, Grahamstown for assisting with identification; Mr. 
Adrien Odgers for useful information and for accompanying me into the re- 
serve on most of my visits; Ms Felicity Kapp and Ms Anita Tait for technical 
assistance. 

References 

Acocks, J. P. H., 1975. Veld types of South Africa. Mem. bot. Surv. S. Afr. 40. 

Dyer, R. A. 1975. The genera of Southern African flowering plants. Vol. 1. Pretoria: 
Department of Agricultural Technical Services. 

, 1976. The genera of Southern African flowering plants. Vol. 2. Pretoria: 

Department of Agricultural Technical Services. 

Wiens, D. and Tolken, H. R., 1979. Viscaceae. In: Leistner, O. A., FI. S. Afr. 
10,1: 43-56. Pretoria: Department of Agricultural Technical Services. 


JlS.Afr. Bot. 49 (2): 175 (1983) 


Short Note 

ALOE MEYERI VAN JAARSVELD, AND A. RICHTERSVELDENSIS 
VENTER & BEUKES 

P. VORSTER 

(Department of Botany, University of Stellenbosch, Stellenbosch 7600 , 
R.S.A.) 

Aloe richtersveldensis Venter & Beukes was published in Kew Bull. 36: 
675 (-678) on 11.6.1982, having been accepted for publication in February 
1981. However, what is obviously the same species, had been described 
under the name Aloe meyeri van Jaarsveld in J1 S. Afr. Bot. 47: 567 (-571) 
on 1.7.1981, having been accepted for publication on 5.3.1981. 

Following Article 11 of the I.C.B.N. (1978 ed.), the correct name for 
this species therefore is Aloe meyeri, and A. richtersveldensis is here reduced 
to a synonym of A. meyeri. 

Venter & Beukes’ paper contains a map showing the distribution of A. 
meyeri and its closest allies, which constitutes an useful addition to van 
Jaarsveld’s paper. 


Accepted for publication 2nd December 1982. 


175 



J1 S. Afr. Bot. 49 (2): 177-180 (1983) 


Book Reviews 

The Biochemistry of Alkaloids, by Trevor Robinson, with pp. viii + 225 
and 35 figures. 2nd ed. New York, Heidelberg, Berlin: Springer- Verlag, 
1981. Volume 3 in the series “Molecular Biology, Biochemistry and Biophy- 
sics”. DM 88, approx. US$ 41,00. ISBN 0-387-10795-9. 

This book is one of the series, Molecular Biology, Biochemistry and Biophysics, 
edited by A. Kleinzeller, G. F. Springer and H. G. Wittmann. It will be of interest 
to plant biochemists, pharmacologists and natural product chemists. 

The contents of this publication have been structured into three major sections. 
The first two chapters are introductory chapters which deal with the definition of al- 
kaloids, their chemical classification based on their carbon-nitrogen skeletons, their 
distribution in nature, their taxonomic significance, the localisation of alkaloids with- 
in the plant, ontogenetic changes in the alkaloid content of the plant, the function of 
alkaloids in the plant and finally a summary of the general theories of alkaloid bio- 
synthesis. The treatment of the major groups of alkaloids then follows with the em- 
phasis on biosynthesis. In recent years one of the most exciting developments in the 
field of alkaloid biosynthesis has been the identification of some of the enzymes 
which regulate the formation of these compounds. This important body of work has 
been incorporated throughout these chapters. The final section deals firstly with the 
metabolism of alkaloids by animals, micro-organisms and higher plants and secondly 
with their pharmacologic and toxicologic effects. 

The book, in hard cover and measuring 170 x 250 mm, is very well written and 
the quality of production is excellent. There are very few errors. The figures and 
tables are adequate although on a number of occasions names of alkaloids are men- 
tioned in the text but no formulae are given. References for each chapter are 
grouped at the end of the chapter and the literature has been covered through to 
June 1980. 

This book can be recommended for postgraduate students and for research work- 
ers. 


W. E. Campbell 


Basidium and Basidiocarp: Evolution, Cytology, Function and Devel- 
opment, edited by Kenneth Wells and Ellinor K. Wells, with pp. xi + 187 
and 117 figures. New York, Heidelberg, Berlin: Springer-Verlag, 1982. DM 
89, approx. US$ 39,60. ISBN 3540-90631-2. 

Another fine book in the Springer Series in Microbiology (Editor: Mortimer P. 
Starr) in which a comprehensive review is given of some recent studies on basidiomy- 
cetes. This major class of the Fungi is examined from multidisciplinary viewpoints 
taxonomic, cytological, molecular biological, physiological and biochemical. A var- 
iety of experimental techniques are employed and special emphasis is placed on the 
basidium and the basidiocarp. 


177 


178 


Journal of South African Botany 


Oberwinkel stresses the importance of comparative morphological studies of the 
basidium and basidiospore as a means of understanding evolution within the Basidio- 
mycotina. McLaughlin reviews the non-nuclear events during basidial and basidio- 
spore ontogeny while Thielke examines recent studies of meiosis in the basidium, es- 
pecially those dealing with the spindle apparatus and associated structures. Lu’s 
studies of meiosis in the basidia of Coprinus cinereus have extended the information 
available on this process to the molecular level. Uno and Ishikawa describe a series 
of studies suggesting that adenosine 3'5 '-cyclic monophosphate is an essential com- 
ponent in the initiation of basidiocarp formation in a form of Coprinus macrorhizus. 
Gruen has brought together the results of his own extensive studies, and those of 
others on the influence of the pileus and somatic mycelium on stipe elongation dur- 
ing the rapid expansion stage. He also reviews and discusses the evidence that the so- 
matic mycelium and/or nutrients enhance stipe elongation and that the numbers of 
hyphal segments increase during this phase. Completing the studies described by 
Gruen is the report by Gooday, who describes his efforts to determine which hyphal 
constituents control stipe elongation by analyzing chemically the major components 
before and after the major period of elongation. 

In addition to the detailed reference list at the end of each chapter, an author in- 
dex and a good subject index are provided. The book is pleasingly printed, has many 
good informative photomicrographs and can be recommended as reference for post- 
graduate courses in Mycology. Most mycologist will find something of value in this 
book. 


Albert Eicker 


Biochemistry and Physiology of Herbicide Action, by C. Fedtke, with pp. 
xi + 202, 43 figures and 58 tables. Berlin, Heidelberg, New York: Springer- 
Verlag, 1982. DM 156, approx. US$ 69,30. ISBN 3-540-1 1231-6. 

Intensive research has been conducted into the properties and characteristics of 
modern herbicides. These substances are of great economic importance. In this book 
Dr. Fedtke clearly shows that they are of very considerable scientific importance as 
well, a fact that is not always adequately appreciated. The nature and scope of the 
book can best be described in this excerpt from the Foreword. “This book describes 
the effects of herbicides on the metabolism of higher plants from the viewpoint of the 
plant physiologist. The material of this book is therefore, as far as possible, divided 
into areas of metabolism. This book intends (1) to present the reader with current 
knowledge and views in the area of herbicide modes of action and (2) to promote the 
future use of herbicides as metabolic inhibitors in plant physiological research to the 
advantage of both, the pesticide and plant science”. 

The book is written in a clear, concise manner, with clear illustrations and tables. 
The structural formulae for the compounds discussed are given. The bibliography is 
excellent and very valuable. 

Until this book appeared, it was difficult to find all this information in one place, 
although some of it was readily available. However, the fact that all this material is 
contained in the one volume makes this book so valuable for plant physiologists. 
They can readily integrate this material into their own studies and it is in an ideal 
form for study by advanced level students and postgraduates. All the colleagues to 
whom I showed this book were very impressed with it and felt it filled a serious gap 
on our shelves. 


Book reviews 


179 


There is too much detailed information in this book to be able to discuss its con- 
tents in detail. But to get an impression of the systematic approach of the author, a 
look at the chapter headings is very useful. Due to lack of space the sub sections of 
the chapters cannot be given, but in Chapter C— Photosynthesis— there are eleven 
sub sections with titles like — Inhibition of Photosynthetic Electron Flow and Physiol- 
ogical Effects Induced by Herbicides that Inhibit Photosynthesis. There are also 436 
references cited in this chapter, which is the longest in the book and takes up half the 
text. 

Contents 

A. Approaches to and Definitions of Mechanisms of Herbicides 

B. Plant Metabolism — a Synopsis of Principles 

C. Photosynthesis 

D. Energy Conversion 

E. Nucleic Acid and Protein Synthesis 

F. Microtubules 

G. Lipid Metabolism 

H. Herbicidal Germination Inhibitors with Unknown Mode of Action 

I. Herbicides with Auxin Activity 

K. Auxin-Inhibitor Herbicides 

L. Aromatic Amino Acid Biosynthesis 

M. Other Herbicides and Mechanisms 

The influence that herbicides had upon nitrogen metabolism was of considerable 
interest to the reviewer as was the matter relating to auxin activity. But the book is 
packed with valuable interesting information. Dr. Fedtke has succeeded in the first 
of his stated aims and his second one, to promote the use of herbicides as metabolic 
inhibitors in research, is sure to follow the reading of this book by experimentalists. 

Highly recommended for plant physiologists, plant biochemists and other plant 
scientists as well as graduate or senior students. 

K. H. Schutte 


Nucleic Acids and Proteins in Plants 1: Structure, Biochemistry and 
Physiology of Proteins, edited by D. Boulter and B. Parthier, with pp. xx 
+ 768 and 135 figures. Berlin, Heidelberg, New York: Springer-Verlag, 
1982. DM 268, approx. US$ 119. ISBN 3-540-11008-9. 

The vast volume of information accumulated on plant physiological processes as 
well as the rapid advances in the technology associated with this field of science over 
the last 15 years necessitated the publishing of a new series of Encyclopedia of Plant 
Physiology. That all libraries should have these books on their shelves is essential. 

The present book, Volume 14A, covers a wide range of topics. These were re- 
viewed by recognised experts in their fields. The editors succeeded in reducing over- 
lap to a level which is essential for clarity and continuity in the text. The 17 chapters 
all contribute to a better understanding of the role of nucleic acids and proteins in 
plants. In order to achieve this understanding the Volume was divided into two sec- 
tions. In the first, topics related to the biosynthesis and metabolism of protein ammo 
acids and proteins were covered. Thus attention was given to ammonia assimilation 
and amino acid metabolism; transfer RNA and aminoacyl-tRNA synthetases in 
plants; ribosomes, polysomes and the translation process; post translational moditica- 


180 


Journal of South African Botany 


tions; protein degradation; physiological aspects of protein turnover; structures of 
proteins; protein types and distribution; cereal storage proteins, structure and role in 
agriculture and food technology; biochemistry and physiology of leaf proteins; micro- 
tubule proteins and P-proteins; plant peptides; and immunology. 

In the second section entitled “Nucleic acids and proteins in relation to specific 
plant physiological processes” attention was given to seed development; protein and 
nucleic acid synthesis during seed germination and early seedling growth; leaf senes- 
cence; and macromolecular aspects of cell wall differentiation. 

Throughout this very neatly presented volume, in which the printing and figures 
are of the highest quality, the authors attempted to state what is currently known and 
how our knowledge has changed over the past two decades. In many cases they iden- 
tified and indicated future areas of research. This should prove most useful for both 
lecturers and students working on proteins and in the general field of nitrogen 
metabolism. This thorough review of the literature ensures that this volume will 
serve as a reference book for many years to come both at the undergraduate and 
postgraduate levels. 


J. van Staden 


J1 S. Afr. Bot. 49 (3): 181-184 (1983) 


CHROMOSOME NUMBERS IN SOME SOUTH AFRICAN SPECIES OF 
LINUM L. (LINACEAE) 

C. M. Rogers 

(Department of Biological Sciences, Wayne State University, Detroit, Michi- 
gan 48202, U.S.A.) 

Abstract 

Chromosome numbers of five (of fourteen) species of section Linopsis of the 
genus Linum of South Africa were counted. 

On gross morphological bases the South African species have been thought to be 
more or less intermediate between species of the Mediterranean region and those of 
the New World. They are characterized by base chromosome numbers of x = 9 (Old 
World) or x = 18 (New World) and it was thought that the chromosome number of 
the South African species would likely be one or the other. The South African 
species examined, however, prove to have a base chromosome number of x = 15 and 
some doubt is raised as to their exact position in the section. 

UlTTREKSEL 

CHROMOSOOMGETALLE IN SOMMIGE SUID-AFRIKAANSE SOORTE 
VAN LINUM L. (LINACEAE) 

Die chromosoomgetalle van vyf (uit veertien) soorte van die seksie Linopsis van 
die genus Linum van Suid-Afrika is bepaal. 

Op 'n uiterlike morfologiese basis is gemeen dat die Suid-Afrikaanse soorte min 
of meer intermedier tussen die soorte van die Middellandse See en die van die Nuwe 
Wereld is. Hulle word gekenmerk deur basiese chromosoomgetalle van x = 9 (Ou 
Wereld) of x = 18 (Nuwe Wereld) en is dit verwag dat die Suid-Afrikaanse soorte 
waarskynlik dieselfde as die een of ander sal wees. Die Suid-Afrikaanse soorte wat 
ondersoek is het egter ’n basiese chromosoomgetal van x = 15 en bestaan daar dus 
nou onsekerheid van hulle presiese plek in die seksie. 

Key words: Linum , Linaceae, South Africa, flax, chromosome numbers. 
Introduction 

The author has been occupied with the classification of the section Li- 
nopsis , of the genus Linum , which includes many of the yellow-flowered 
species of both Old and New Worlds. Among other things, the chromosome 
numbers of many of the species, except those of South Africa, have become 
known. 

The section ranges from southern Europe to eastern and southern Africa 
and to North and South America, with a single outlier in India and Sri Lar> 


Accepted for publication 25th November, 1982. 


181 


182 


Journal of South African Botany 


ka. The species of the Mediterranean are mostly diploid, with n = 9 or 10. 
One small group, the subsection Dichrolinum , which includes L. suffrutico- 
sum L., L. tenuifolium L. and their allies, includes taxa with n = 9, 18 and 
36. The small subsection Halolinum , with L. trigynum L. and the heterosty- 
lous species L. tenue Desf. and L. maritimum L., is characterized by n = 10. 
They differ from others in the section in having linear stigmas and thus ap- 
pear to be out of the main line of evolution in the section, other species of 
which have capitate stigmas. The latter thus far have all been found to have 
x — 9. Mediterranean species, such as L. setaceum Brot., and L. strictum L., 
for example, have n = 9, while L. volkensii Engl., of eastern Africa, is in- 
terpreted as a hexaploid with n = 27 (Lewis, 1964). In South America the 
eight species (of fourteen) thus far examined have been found to have n = 
18 or 36 (Mildner and Rogers, 1978), while the base number for the numer- 
ous North American species of the section also proves to be x = 18 (Rogers, 
1969). 

The South African species are morphologically very similar to species of 
the New World and, therefore, they were expected to follow the same pat- 
tern and it would mainly be a matter of discovering whether they might still 
retain the primitive diploid number of n = 9 or the tetraploid number of n = 
18 of the apparently closely allied New World species. Earlier reports (in 
Darlington and Wylie, 1955) of L. africanum having n = 15 or 16 were not 
given much credence, since seed of African-grown plants of L. usitatissimum 
L., the cultivated flax, which is said to have these chromosome numbers, is 
frequently distributed under the name “L. africanum" . 


Material and Methods 

In the early part of 1982, the author had the opportunity to collect some 
material of several South African species, all from the Cape Province. They 
included Linum acuticarpum, L. africanum , L. comptonii, L. gracile and 
L. heterostylum. Flower buds were collected and preserved in Carnoy’s 
fluid. Anthers were later squashed in acetocarmine and chromosomes of first 
or second meiotic division examined. 

Voucher specimens are deposited in the herbarium of Wayne State Uni- 
versity (WUD); duplicates are presently being prepared for distribution to 
the major herbaria of South Africa. 


Results 

Chromosomes of all the species of section Linopsis examined thus far 
(approximately fifty species) have proved to be very small, averaging about 
2 /x in length at first metaphase of meiosis. Those of South African species 


Chromosome numbers in species of Linum L. (Linaceae) 183 

prove to be no exception and so no attempt has been made to discern details 
of chromosome structure and only numbers of chromosomes were determined. 
The following results were obtained: 

Linum acuticarpum Rogers n = 15. Roadside, ca 1 km west of the 
bridge over the Palmiet R., along R 44, Caledon District. Rogers 13704, 
19 Feb., 1982. 

Linum africanum L. n = 29. Roadside, Heuningkloof, near Gordon’s 
Bay, Somerset West District. Rogers 13683 , 1 Feb., 1982. 

Linum comptonii Rogers n = 15. Moist soil, Hangklip, Caledon Dis- 
trict. Rogers 13693, 10 Feb., 1982. 

Linum gracile Planch, n = 15. Half way up the south slope of Outeni- 
quas Pass, north of George, George District. Rogers 13696 , 13 Feb., 1982. 

Linum heterostylum Rogers n = 15. (1) Geelrug on Rietfontein Private 
Nature Reserve. Bredasdorp District. Rogers 13684, 2 Feb., 1982. (2) Near 
De Hoop Provincial Farm, near Bredasdorp, Bredasdorp District. Rogers 
13685, 3 Feb., 1982. (3) Slopes, 2-3 km from Barrydale toward Tradouw 
Pass, Swellendam District. Rogers 13694, 11 Feb., 1982. 


Discussion and Conclusions 

Clearly the South African species, at least insofar as chromosome counts 
have now been made (there are nine additional species which have not been 
examined), have a base number of x = 15 and are not direct intermediates 
between the Mediterranean and the New World species of the section. 
Linum africanum is a variable species, of which the plants examined are 
somewhat different from the Linnaean type. The chromosome number of 
the population sampled is almost certainly an aneuploid derivative of the 
number n = 30. 

Very possibly Linum mysurense Heyne ex Benth., of India and Sri 
Lanka, is related most closely to South African species. It is rather distantly 
disjunct from any other species of the section, except for species such as 
Linum strictum L. and L. trigynum L., which in recent times have become 
widely adventive, and it has been distinguished from all other Old World 
species of the section in having a chromosome number of n = 30 (Rogers, 
1975). Until the South African species were examined no other Old World 
species of the section were known which might be derived from a base num- 
ber of x = 15. 

Examination of additional species, especially L. thunbergii Eckl. & 
Zeyh. which links, geographically and morphologically, species of east Afri- 
ca with those of South Africa, will very possibly throw some further light on 
the relationships of the latter to others on the continent as well as to those 
of the New World. 


184 


Journal of South African Botany 


References 

Darlington, C. D. and Wylie, A. P., 1955. Chromosome Atlas of Flowering Plants. 
London: Geo. Allen and Unwin. 

Lewis, W. H., 1964. A hexaploid Linum (Linaceae) from eastern Ethiopia. Sida 1 : 
383-384. 

Mildner, R. A. and Rogers, C. M., 1978. Revision of the native South American 
species of Linum (Linaceae). Phytologia 39: 343-390. 

Rogers, C. M., 1969. Relationships of the North American species of Linum (flax). 
Bull. Torr. bot. Club 96 : 176-190. 

,1975. In: 10PB chromosome number reports XLIX. Taxon 24 : 503. 


J1 S. Afr. Bot. 49 (3): 185-188 (1983) 


GALIUM ROURKEI (RUBIACEAE): A NEW SPECIES FROM THE 
SOUTH WESTERN CAPE 

Ch. Puff 

( Institute of Botany, University of Vienna, Rennweg 14, A-1030 Vienna, 
Austria) 

Abstract 

Galium rourkei, a new species from the south western Cape, is described. Its af- 
finities to the two Cape endemic species G. subvillosum and G. mucroniferum are 
discussed, and the isolated position of this group among the southern African Galium 
species is pointed out. 

UlTTREKSEL 

GALIUM ROURKEI (RUBIACEAE): N NUWE SOORT VANAF DIE SUID- 
WESTELIKE-KAAP 

Galium rourkei, ’n nuwe soort vanaf die suidwestelike-Kaap, word beskryf. Die 
soort se affiniteite met die twee Kaapse endemiese soorte G. subvillosum en 
G. mucroniferum word bespreek en die gei'soleerde posisie van die groep tussen die 
endemiese soorte van Galium van suidelike Afrika word aangetoon. 

Key words: Galium, sp. nov., Rubiaceae, south western Cape. 

Dr. John Rourke has kindly drawn my attention to an apparently very 
rare and as yet undescribed Galium species from the Kogelberg Forest Re- 
serve in the south western Cape. 

Galium rourkei Puff, sp. nov. similis G. subvillosi Sond. sed cymis 2-floris 
et foliis stipulisque foliaceis minoribus 4-5-natim verticillatis differt; affinis 
G. mucroniferi Sond. sed foliis oblanceolatis vel spathulatis multo latioribus 
facile distinguenda. 

Herba perennis, prostrata, ramosissima. Caules graciles, ± quadrangu- 
lares, c. 100-200 mm longi. Folia et stipulae foliaceae 4-5-natim (rarissime 
6-natim) verticillata, 5-7 mm longa, (2)2, 5-5 mm lata, oblanceolata vel spa- 
thulata, pilosa, apicibus cuspidatis-mucronatis. Cymae 2-florae, pedunculi 
1,5-5 mm longi, pedicelli 1,5-3 mm longi sub anthesi, postfloraliter usque 
ad 6 mm elongati, divaricati. Corolla cremeo-alba, rotata, (2, 5)3-3, 5(4) mm 
diam., lobi 4, ± ovati. Fructus niger, pilis albis obtectus; mericarpium unum 
saepe abortum. 


Accepted for publication 23rd December, 1982. 


185 


186 


Journal of South African Botany 



Fig. 1. 

Galium rourkei. a, a node from the midstem region: 5 leaves in a whorl; b, bud; c, 
flower from above. Camera lucida drawings from the type collection; the bar re- 
presents 1 mm. 


Type: South Africa, Cape Province — 3418 (Simonstown): Kogelberg Forest Re- 
serve, Kuduberg (-BD), c. 830 m, 17.12.1981, Rourke 1765 (NBG, holo.; WU, iso). 

Perennial. Stems ± much-branched, delicate, ± prostrate, c. 100- 
200 mm long and c. 0,5-1 mm in diam., ± distinctly 4-angled, at least 
younger parts with curved or straight white hairs on the angles, c. 0,4-0, 7 
mm long. Longest internodes c. 8-12(17) mm. Middle cauline leaves (Fig. 
la) in whorls of 4-5(-6), 1-nerved, 5-7 x (2)2, 5-5 mm, oblanceolate to spa- 
thulate, with a distinct, rigid point at the apex, up to 1 mm long; with 
straight or curved hairs, c. 0,3-0, 6 mm long on the upper and lower surface 
and margins; margins straight. Synflorescences much-reduced; partial inflo- 
rescences 2-flowered; peduncles 1,5-5 mm long, hairy; pedicels filiform, 
hairy, 1,5-3 mm at anthesis, to 6 mm and divaricate in fruit, only rarely sub- 
tended by lanceolate bracts, c. 1,5 x 0,5 mm. Flowers (Fig. lc) strongly pro- 
tandrous, 4-merous; corolla (2, 5)3-3, 5(4) mm in diam., rotate, creamy- 
white; lobes ± ovate, longer than wide, acute, with a few white, straight or 
curved hairs on the margins; stamens c. 1-1,5 mm long, filaments filiform, 
anthers brownish, ± ellipsoidal; styles very short, 0,5 mm or less, stigma ± 
capitate. Buds (Fig. lb) ± ovoidal. Fruits dry, blackish, covered with white 
spreading hairs, c. 0,3-0, 5 mm long; mericarps ± globose, each c. 1,5-2 
mm in diam.; often only one mericarp developed. 

Average pollen diameter: 27,63 /im. 

Habitat; In permanent shade under damp rock cliffs, in south-facing situ- 
ation. Growing in association with mosses and Hymenophyllum (notes from 
the type collection). 


Galium rourkei (Rubiaceae): a new species 


187 


Although only known from the type collection, I have no doubt that 
G. rourkei is a “good” new species. It is allied to both G. subvillosum and 
G. mucroniferum . primarily on account of the similar, reduced inflorescences 
(3-or 2-flowered partial inflorescences) and flowers with stamens (filaments) 
almost as long as the corolla lobes. Also similar habitat conditions (shel- 
tered, cool, moist) and corresponding average pollen diameters, larger than 
those of other Cape species of Galium (except for G. undulation , an unlikely 
ally which markedly differs in floral and inflorescence structure), suggest an 
affinity. The large average pollen diameter may, furthermore, indicate poly- 
(?tetra)ploidy, which has been documented for both G. subvillosum and G. 
mucroniferum (Puff, 1978). 

G. rourkei differs from both species in having its leaves arranged in 
whorls of 4 or 5 (6 leaves per whorl are the exception), shorter and broader 
leaves (leaf size and shape of some forms of G. subvillosum may ± overlap 
with those of G. rourkei, but the former has always 3- rather than 2- 
flowered partial inflorescences), much more delicate stems and usually 
shorter internodes. 

It is noteworthy that this group of three Cape species seems quite iso- 
lated and does not appear to have any close allies in either the Cape or any- 
where else in southern Africa, much in contrast to, for example, the G. 
capense group (G. capense and the allied Cape species G. bredasdorpense, 
G. monticolum and G. amatymbicum). The latter is allied to G. bussei in 
south central and tropical east Africa, G. ossirwaense , to an as yet undes- 
cribed species from Ethiopia and, ultimately, to the Galium verum alliance 
in Europe and Asia. In an article dealing with the evolution of the African 
and Madagascan Galium species (in preparation), these two contrasting 
groups of Galium Cape species will be discussed in detail. 

To accommodate G. rourkei in the key to the southern African species 
of Galium (Puff, 1978: 220), the beginning of the key needs to be altered as 
follows: 


When this species was described (Puff, 1978: 244), only the type specimen was 
known. In the meantime, the following additional collections have been brought 

to mv attention: , , , in , n,,roerK 

CAPE-3420 (Bredasdorp): De Hoop near T.erhoek, (-AD), 3°m, Burgm 

2133 (STE), v. d. Merwe 2001 (STE); farm Windhoek, v. d. ' 20i ( 
Fonteinkloof on W side of De Hoop vlei, 120 m. Burgers 1-51 (5> 1 E T 
These collections, furthermore, confirm my original assumption that^the 
species may be confined to limestone areas. The listed specime „ g 
cracks and holes in limestone. 


l'88 


Journal of South African Botany 


1. Leaves 3-nerved, strictly in whorls of 4 1. G. thunbergianum 

Leaves 1-nerved, in whorls of 4-10 la 

1A. Leaves in whorls of 4-5 (rarely 6), 5-7 x (2)2, 5-5 mm, oblanceolate to 

spathulate 10A. G. rourkei 

Leaves in whorls of 6-10, much larger or, if small, ± ovate-lanceolate or linear- 
lanceolate 2 

2. Synflorescences 2 

References 

Puff, C., 1978. The genus Galium L. (Rubiaceae) in southern Africa. Jl S. Afr. Bot. 

44 : 203-279. 


J1 S. Afr. Bot. 49 (3): 189-192 (1983) 


A NEW EUPHORBIA FROM SOUTH WEST AFRICA 

L. C. Leach 

( Botanical Research Institute , Private BagXIOl, Pretoria , R.S.A.) 

Abstract 

A new tuberous rooted dwarf Euphorbia with a very restricted distribution on 
the Namuskluft, near Rosh Pinah, just north of the Orange River, is described. 
Vegetatively indistinguishable from E. wilmaniae Marl., its floral and fruiting charac- 
teristics are entirely different. 

UlTTREKSEL 

'N NUWE EUPHORBIA VANAF SUIDWES-AFRIKA 

’n Nuwe dwerg Euphorbia met ’n wortelknol en met 'n baie beperkte versprei- 
ding op die Namuskluft, naby Rosh Pinah, net noord van die Oranjerivier, word be- 
skryf. Vegetatief nie te onderskei van E. wilmaniae Marl, nie, maar blom- en vrug- 
kenmerke verskil heeltemal. 

Key words: Euphorbia, sp. nov., Euphorbiaceae, South West Africa/Namibia, Ga- 
riep Centre. 

Euphorbia namuskluftensis Leach, sp. nov. ad E. wilmaniae Marl, in 
characteribus vegetativis simillissima sed ilia glandulis non prominenter ru- 
gosis sine processibus dactylinis; ovulo vix exserto seminibusque minoribus 
proportione angustioribus statim distinguenda. 

Planta dioecia, succulenta, inermis, ramosissima; ramis ramulisque aeriis 
confertis, raro in longitudinem 20 mm attingentibus; radice tuberosa rhizo- 
matosa, solum in terra calcarea crescens. Rami ramulique tuberculati, esti- 
pulares micro-scabrelli, c. 5 mm crassi. Folia carnosa, ovata, acuta, leviter 
plicata, patulo-recurva, variabiliter micro-scabrella, decidua, cicatrice brun- 
nea tuberculi ad apicem remanenti. Inflorescentia axillaris, ad apicem ramo- 
rum, plerumque cyathio solitario unisexuali in pedunculo bracteato portato. 
Bracteae plus minusive oblongae, aliquanto obtusae, irregulariter denticula- 
tae, plerumque apicula breviter plicata, acuta, recurva instructae, anguste 
subpellucido-marginatae, saepe in superficie sparsim pubescentes. Cyathium 
glabrum vel raro sparsissime micro-scabrellum, 2-3 mm longum, c. 3 mm 
diam.; glandulae 5, erectae, aliquanto cuneatae, 1-2,25 mm latae, parum bi- 
labiatae, labio interiore parvulo, infra laeves, atro-virides, supra palles- 
centes, perleniter rugulosae, recurvae; lobi 5, subquadrati, c. 1 mm longi, ir- 


Accepted for publication 13th January, 1983. 

189 


190 


Journal of South African Botany 


regulariter fimbriati, utrinque pilosi. Involucrum , femineum intus glabrum, 
apice piloso excepto; florum, septorum, bracteolarumque vestigiis instruc- 
tum. Flos femineus , pedicello sparsim minute puberulo, c. 0,5 mm longo suf- 
fultus; ovarium subglobosum subsessile; stylis crassis, in columnam basi con- 
natis, partibus libris patulo-recurvis, ad apicem breviter bifidis, lobis patulis 
valde recurvis. Involucrum masculum intus versus apicem pilosum, septis 5; 
septa ad medium in lobis duobus plumosis bracteoliformibus divisa. Flores 
masculi c. 40, aliquot bracteolis plumosis, c. 2 mm longis; pedicelli pubes- 
centes, albidi; filamenta pallide viridia, minute puberula, c. 0,75 mm longa; 
antheris atro-viridibus, polline flavido-auriantiaco. Capsula glabra, tantum 
ex involucri reliquiis exserta, subglobosa, leviter 3-sulcata, atro-brunnea, c. 
5,75 mm diam., 4,75 mm alta. Semen late suboviodeum, leviter ruminato- 
glebulosum, brunneo-cremeum, c. 3,5 mm x 2,5 mm. 

Typus: S.W.A., Namuskluft, Lavranos & Pehlemann 20796 (PRE, holo; 
WIND). 

Plant dioecious, much-branched, unarmed, succulent, rhizomatose- 
tuberous rooted, with the above ground crowded branches and branchlets 
seldom attaining 20 mm in height (plants in cultivation may develop 
branches up to 60 mm or more long, while rhizome-like underground 
branches are often longer). Branches and branchlets tuberculate, estipulate, 
micro-scabridulous, ± 5 mm thick. Leaves ovate acute, ± 1,5 mm long, 
spreading recurved, somewhat folded, fleshy, variably micro-scabridulous, 
deciduous, leaving a brown scar at the apex of the tubercle. Inflorescence ax- 
illary, at the apex of the branches, usually a single unisexual cyathium borne 
on a short, glabrous, bracteate peduncle with a pseudo-whorl of three bracts 
below the cyathium. Bracts quite different from the leaves, especially in the 
male plant, ± oblong, somewhat obtuse, irregularly denticulate, with a nar- 
row subpellucid border, usually with a shortly folded, recurved, acute apicula, 
often somewhat pubescent on the upper surface. Cyathium glabrous or rare- 
ly very sparsely micro-scabridulous, green at the base becoming purplish 
above, 2-3 mm long, ± 3 mm in diam.; glands 5, erect, somewhat cuneate, 
1-2,25 mm broad, slightly two-lipped, with a thickened rim and a small cen- 
tral lip on the inner margin, slightly concave, dark green, smooth below, be- 
coming pale on the very faintly rugulose, recurved outer margin; lobes 5, 
subquadrate, ± 1 mm long, irregularly fimbriate, pilose both sides, more 
densely within. Involucre $, glabrous inside, except towards the apex, there 
pilose, particularly densely below the gland, with the rudiments of septa, 8 
flowers and bracteoles around the base (unusual in $ cyathia of unisexual 
species); $ flower, almost subsessile, pedicel sparsely minutely puberulous, 
± 0,5 mm long; perianth rim-like; ovary subglobose; styles very stout, united 
at the base into a short stout column, ± 0,6 mm thick, ± 


New Euphorbia from South West Africa 


191 


0,5 mm long, the free portions spreading recurved. Male flowers, ± 40, with 
a few plumose bracteoles, ± 2 mm long; pedicels ± 2 mm long, pubescent, 
whitish; filaments ± 0,75 mm long, pale green, minutely puberulous; anther 
thecae dark green with yellow-orange pollen. Capsule glabrous, barely ex- 
serted from the remains of the cyathium, subglobose, shallowly 3-grooved, 
dark brown, ± 5,75 mm diam., 4,75 mm high. Seed broadly sub-ovoid, shal- 
lowly ruminate-glebulose, brownish cream. ± 3,5 mm x 2,5 mm. 

Material Examined 

SOUTH WEST AFRICA— 2716 (Witputz): Namuskluft (-DD), 1 200 m alt., re- 
stricted to white limestone outcrop, st. Oct. 1978, Lavranos & Newton 16871 (K), 
cult. PRE, ? Sept. 1982, Lavranos & Newton 17167 (PRE, WIND), Lavranos & 
Pehlemann 20796 6 (PRE. WIND), 20797 $ , fr. (PRE). 

Euphorbia namuskluftensis is virtually indistinguishable from E. wilman- 
iae in habit and vegetative characters but differs widely in its floral, with 



Fig. 1. 


Euphorbia namuskluftensis Leach, sp. nov. 1. Male cyathia, 2. Female cyathia, 3. 
Peduncular bracts; 4. Fascicle of male flowers; 5. Bracteole, 6. Lobe, 
tened); 8. Female flower; 9. Female involucre. 


192 


Journal of South African Botany 



Fig. 2. 

Euphorbia namuskluftensis Leach, sp. nov. 10. Mature capsule; 11. Perianth; 12. 

Seed; 13. Leaf. 


smooth (not prominently rugose) glands which lack any sign of the charac- 
teristic finger-like processes of that species, and a subsessile capsule with 
smaller proportionally narrower seeds. 

This species has been found only on a white limestone outcrop on the 
Namuskluft, where it is associated with the recently described E. lavrani 
Leach. As with that species, the only plant in cultivation at PRE was fe- 
male, and although clearly quite different from E. wilmaniae it was consid- 
ered that the material available was inadequate for its publication as a new 
species. I am consequently most grateful to John Lavranos for returning to 
the locality and collecting a male plant with adequate mature material. 


J1 S. Afr. Bot. 49 (3): 193-212 (1983) 


A CHECK-LIST OF THE PTERIDOPHYTES OF ZIMBABWE 

J. E. and S. M. Burrows* 

( Sengwa Wildlife Research Institute, Private Bag 6002, Gokwe, Zimbabwe) 
Abstract 

The two hundred and thirty-four species of ferns and fern allies indigenous or 
naturalised in Zimbabwe are listed, thereby up-dating E. A. Schelpe's (1970) treat- 
ment by including new records and name changes effected since his publication. 

UlTTREKSEL 

’N KONTROLELYS VAN DIE PTERIDOFIETE VAN ZIMBABWE 

Die tweehonderd vier-en-dertig varingsoorte en varingagtiges, inheems of gena- 
turaliseerd in Zimbabwe word aangegee en so word E. A. Schelpe se (1970) verhan- 
deling op datum gebring deur die insluiting van nuwe varingsoorte en naamveran- 
derings sedert sy publikasie. 

Key words: Pteridophytes, Zimbabwe, ferns. 

Introduction 

The following is a check-list of the ferns and fern allies occurring in Zim- 
babwe. Since the publication of Schelpe’s Pteridophyta for the Flora Zambe- 
siaca (1970), there have been several new records, new taxa described and 
several name changes making an up-dated check-list desirable. 

The arrangement and numbering of families and genera follows Crabbe, 
Jermy and Mickel’s numerical sequence (1975). Although differing radically 
from Schelpe’s treatment, this is now a more widely accepted arrangement. 
Schelpe’s basic generic concept has nevertheless been followed, with a few 
exceptions. 

Synonyms have been printed in italics, and since much of the origin of 
Zimbabwean synonymy dates from T. R. Sim’s Ferns of South Africa 
(1915), synonyms and misapplied names from this source are also provided 
in the list. 

References to recent generic or family revisions since the Flora Zambe- 
siaca volume on Pteridophyta was published are given under the relevant 
taxa. 


Present address: Coromandel Farm, Private Bag X1017, Lydenburg 1120, R.S.A. 
Accepted for publication 17th January, 1983. 


193 


194 


Journal of South African Botany 


Distribution of species in surrounding territories is indicated in the list by 
numerals, as follows: 

0 Endemic to Zimbabwe or so far only recorded from Zimbabwe 

1 Botswana 

2 Zambia 

3 Malawi 

4 Mozambique 

5 Area covered by the flora of southern Africa, but excluding Botswana 
(i.e. Republic of South Africa, Transkei, Lesotho, Swaziland, Bophu- 
thatswana, Namibia and the Caprivi Strip). 

Where species do not occur in any of the countries adjacent to Zim- 
babwe, but are found elsewhere in Africa or the Indian Ocean islands, the 
nearest locality to Zimbabwe is listed. 

A specimen is cited for each taxon. Where possible specimens have been 
cited which are housed in the National Herbarium of Zimbabwe (SRGH), 
with duplicates in two or more of the following herbaria: K, BOL, BM, 
PRE, LISC. Where a species is described from Zimbabwe the specimen 
quoted is the type and the collector's name italicised. 

Introduced species which have now become naturalised in parts of the 
country are indicated by an asterisk. Cultivated plants are not included in 
the list. 

The habitat or life-form is indicated as follows: 

E epiphytic 

L lithophytic 

T terrestrial 

A aquatic 

C climber 

Where a species is able to grow in more than one habitat, more than one 
letter is given, e.g. ETL. The term lithophytic is here given to plants which 
adhere to a rock surface, whether bare or covered with a layer of moss or 
lichen. Where a plant grows in a deposit of soil or humic matter on top of, 
or among rocks, it has been classified as terrestrial. Aquatics are either free- 
floating or rooted in water for all or part of the year, but the term does not 
include waterside plants. 

Distribution within Zimbabwe is indicated by letters which show the 
geographical divisions as used in the Flora Zambesiaca: N, W, C, E, S (see 
Fig. 1). 

Acknowledgements 

The assistance of the following persons is gratefully acknowledged; Pro- 
fessor E. A. Schelpe and Mrs. N. C. Anthony of the Bolus Herbarium, 


Check-list of pteridophytes of Zimbabwe 


195 


Cape Town; Mr. R. B. Drummond and Mr. G. V. Pope of the National 
Herbarium. Harare. 



Fig. 1. 

The geographical divisions of Zimbabwe. 


001 Psilotum Sw. 

P. nudum (L.) Beauv. 


PSILOTACEAE 

245 NWCES Wild 6601 


LE 


LYCOPODIACEAE 


003 Lycopodium L. 

L. carolinianum L. 

var. affine (Bory) Schelpe 234 

var. grandifolium Spring 2345 
L. cernuum L. 2345 

L. clavatum L. 

var. inflexum (Beauv.) 

Spring 2345 

L. dacrydioides Bak. 345 


E Goodier & Phipps 229 T 

WES Robinson 1932 T 

NWCES Miller 5895 T 


T 

E 


E 

E 


Chase 4507 
Muller 3407 


196 


Journal of South African Botany 


Lycopodium L. 


L. gnidioides L.f. 

345 

E 

Goodier 170 

LE 

L. ophioglossoides Lam. 

345 

E 

Chase 3182 

LE 

L. saururus Lam. 

35 

E 

W. Jacobsen 3745 

LT 

L. verticillatum L.f. 

2345 

E 

Mitchell 502 

LE 


SELAGINELLACEAE 


008 Selaginella Beauv. 


S. abyssinica Spring 

234 

NE 

Chase 5041 

LT 

S. caffrorum (Milde) 
Hieron. 

5 

S 

J. Burrows 2691 

L 

S. dregei (C.Presl) Hieron. 

12345 

NWCES Chase 5188 

TL 

S. imbricata (Forsk.) Spring 
ex Decne. 

245 

NW 

Chase 5763 

T 

S. kraussiana (Kunze) 
A. Braun 

345 

E 

Wild 3563 

T 

S. mittenii Bak. 

2345 

NWCES 

Pope 126 

T 

S. nivea Alston 

4 

S 

Leach & Cannell 14894 

T 

S. perpusilla Bak. 

Tanzania S 

Drummond 7851 

T 


ISOETACEAE 


009 


Isoetes L. 

I. aequinoctialis Welw. ex 
A. Braun 

I. schweinfurthii A. Braun 


25 

125 


I. alstonii Reed & Verdcourt 
I. rhodesiana Alston 


N Whellan 849 A 

WS Drummond & 

Rutherford-Smith 7557 
& 7557a A 

Greenway & Brenan 
8012 

West 3075 


Oil Equisetum L. 

E. ramosissimum Desf. 


EQUISETACEAE 

2345 NWCES Phipps 864 


T 


OPHIOGLOSSACEAE 


016 Ophioglossum L. 


O. costatum R.Br. 

234 

NWCES 

Drummond 5452 

T 

O. gomezianum Welw. ex 
A. Braun 

235 

NCES 

Drummond 5102 

T 

O. lancifolium C.Presl 

23 

NWE 

Mitchell 406 

T 

O. polyphyllum A. Braun 

1345 

NCES 

Phipps 1463 

T 

O. reticulatum L. 

2345 

NWCES 

J. Burrows 2083 

T 

O. vulgatum L. 

2345 

NCE 

Chase 5913 

T 


O. vulgatum var. 
kilimandscharicum 
Hieron. 


Check-list of pteridophytes of Zimbabwe 197 


MARATTIACEAE 

022 Marattia Sw. 

M. fraxinea Sm. ex 
J.F.Gmel. 

var. salicifolia (Schrad.) 



C.Chr. 

2345 

ES 

Chase 4508 

T 



OSMUNDACEAE 



026 

Osmunda L. 






O. regalis L. 

2345 

NWCES 

Plowes 1739 

T 

029 

Todea Willd. 






T. barbara (L.) Moore 

45 

E 

Mitchell 278 

T 



SCHIZAEACEAE 



032 

Schizaea Sm. 






S. pectinata (L.) Sw. 

345 

E 

J. Burrows 2440 

T 


S. tenella Kaulf. 

5 

E 

J. Burrows 2410 

T 

034 

Lygodium Kuhn 






L. kerstenii Kuhn 

245 

ES 

Chase 6975 

TC 


L. brycei Bak. 



Bryce s.n. 


035 

Anemia Sw. 






A. angolensis Alston 
A. simii var. angustiloba 

23 

N 

Phipps 956 

T 


(Bonap.) P.-Sermolli 
A. simii Tardieu 
A. anthriscifolia sensu Sim 

45 

NCES 

Eyles 726 

T 

036 

Mohria Sw. 






M. caffrorum (L.) Desv. 

2345 

NCES 

Chase 7033 

T 


M. lepigera (Bak.) Bak. 

234 

WCES 

J. Burrows 2577 

T 


PARKERIACEAE 

037 Ceratopteris Brongn. 

C. cornuta (Beauv.) Le 

Prieur 124 NW Guy 1866 A 

C. thalictroides sensu 
Schelpe 


ADI ANT ACE AE 


039 Actiniopteris Link 

A. dimorpha P.-Sermolli 

2345 

NWCES 

Rodin 4407 

LT 

A. pauciloba P.-Sermolli 

2 

N 

Mitchell 560 

LT 

A. radiata (Sw.) Link 

125 

NWES 

Phipps 919 

LT 


8J — 2 


198 


Journal of South African Botany 


047 Cheilanthes Sw. 


Anthony, N. C. & Schelpe, E. A. C. 

L. E. 

in Contr. Bol. Herb. 10: 143— 

164 (1982) 

C. bergiana Schlechtend. 
C. concolor (Langsd. & 
Fisch.) Schelpe & 

345 

E Chase 5990 

T 

N.C. Anthony 
Doryopteris concolor 
(Langsd. & Fisch.) Kuhn 

12345 

NWCES Chase 6802 

LT 

C. eckloniana (Kunze) Mett. 
Notholaena eckloniana 
Kunze 

5 

CES Rushworth 803 

T 

C. farinosa (Forsk.) Kaulf. 

23 

WCE Schelpe 3930 

ET 

C. hirta Sw. 

C. inaequalis (Kunze) Mett. 

145 

NWCES Rodin 4351 

T 

var. inaequalis 

Notholaena inaequalis 
Kunze 

var. buchananii (Bak.) 

235 

NWCES Fisher & Schweickerdt 

476 LT 

Schelpe 

Notholaena buchananii Bak. 

245 

CES Chase 7739 

LT 

C. leachii (Schelpe) Schelpe 
Notholaena leachii Schelpe 
C. multifida (Sw.) Sw. 

234 

WCES Pope 1108 

T 

subsp. multifida 
subsp. lacerata 
N.C. Anthony & 

5 

S J. Burrows 39 

T 

Schelpe 

C. pentagona Schelpe & 

2345 

NWCES Chase 6404 

T 

N.C. Anthony 
C. quadripinnata (Forsk.) 

5 

N Mitchell 105 

T 

Kuhn 

Pellaea quadripinnata 
(Forsk.) Prantl 
C. viridis (Forsk.) Sw. 

345 

NCE Wild 1397 

T 

var. viridis 

Pellaea viridis (Forsk.) 

Prantl 

var. viridis 

var. glauca (Sim) Schelpe 

345 

WCES Wild 2187 

T 

& N.C. Anthony 
Pellaea viridis (Forsk.) 
Prantl 

var. glauca (Sim) Sim 
var. obscura 

12345 

NWCES J. Burrows 2654 

T 

N.C. Anthony 
Pellaea viridis (Forsk.) 
Prantl 

var. involuta sensu 

45 

NWCES H. Burrows 1942 

T 


Schelpe 


Check-list of pteridophytes of Zimbabwe 


199 


Cheilanthes welwitschii Hook, ex 

Bak. 4 NC Mitchell 477 T 

C. mossambicensis Schelpe 

050 Aspidotis (Nutt, ex Hook.) 

Copel. 

A. schimperi (Kunze) 



P.-Sermolli 

234 

NCE 

Chase 4747 

T 

057 

Hypolepis schimperi 
(Kunze) Hook. 

Pellaea Link 
P. boivinii Hook. 

25 

NWCES 

Chase 6163 

T 


P. calomelanos (Sw.) Link 
var. calomelanos 

12345 

NWCES 

Miller 2786 

T 


Pellaea hastata sensu Sim 
var. swynnertoniana 
(Sim) Schelpe 

34 

ES 

Schelpe 5393 

T 


P. swynnertoniana Sim 
P. doniana J.Sm. ex Hook. 

234 

NES 

Swynnerton 850 
Chase 7491 

T 


P. dura (Willd.) Hook. 

2345 

NWCES 

Chase 6230 

T 


P. longipilosa Bonap. 

234 

NC 

Rodin 4427 

T 


P. pectiniformis Bak. 

2345 

NCES 

Chase 3494 

LT 

062 

P. goudotii C.Chr. 

Anogramma Link 
A. leptophylla (L.) Link 

5 

NE 

Eyles 563 

T 

063 

Pityrogramma Link 
P. argentea (Willd.) Domin 

2345 

NCES 

Schelpe 5662 

LT 


Gvmnogramma argentea 
'(Willd.) Mett. 

086 AdiantumL.. 


A. capillus-veneris L. 

2345 

NWCES 

Greenway & Brenan 

ELT 

T 

A. hispidulum Sw. 

34(5*) 

E 

8027 

J. Burrows 2160 

A. incisum Forsk. 

2345 

NWCES 

Schelpe 4084 

LT 

A. caudatum sensu Sim 
A. mendoncae Alston 

4 

NES 

Mitchell 404 

T 

A. patens Willd. 

subsp. oatesii (Bak.) 
Schelpe 

2 

W 

J. Burrows 2630 

T 

A. oatesii Bak. 

A. philippense L. 

2345 

NWCE 

Oates s.n. 
Rogers 5549 

LT 

A. lunulatum Burm.f. 
A. poiretii Wikstr. 
var. poiretti 

2345 

NWCES 

Eyles 328 

LT 

A. thalictroides Willd. ex 
Schlechtend. 

var. sulphureum (Kaulf.) 
Try on 

35 

NCE 

Chase 6536 

LT 

A. raddianum C.Presl 

345 

E 

Chase 6343 

T 


200 


Journal of South African Botany 


093 Vittaria Sm. 

V. elongata Sw. 

4 

E 

J. Burrows 2761 

E 

V. guineensis Desv. 

var. orientalis Hieron. 

234 

E 

Chase 6641 

TLE 

V. isoetifolia Bory 

345 

E 

Fisher 1571 

LE 

V. volkensii Hieron. 

34 

E 

Mitchell 268 

LE 

096 Pteris L. 

Schelpe, E.A.C.L.E., in Contr. Bol. 

Herb. 1: 

57-67 (1969) 


P. buchananii Bak. ex Sim 

5 

E 

Wild 2216 

T 

P. cretica L. 

235 

CES 

Chase 3525 

T 

P. dentata Forsk. 

2345 

E 

Chase 6240 

T 

P. dentata Forsk. 

subsp ./label lata (Thunb.) 
Runemark 

P. pteridioides (Hook.) 
Ballard 

34 

E 

Wild 2199 

T 

P. quadriaurita Retz. 

subsp. catoptera (Kunze) 
Schelpe 

2345 

NE 

Wild 2204 

T 

P. catoptera Kunze 

var. horridula Schelpe 

34 

E 

Chase 4890 

T 

subsp. friesii (Hieron.) 
Schelpe 

2345 

WCES Mitchell 563 

T 

P. friesii Hieron 
P. vittata L. 

2345 

NWCES Gilliland 203 

T 

P. mildbraedii Hieron. 

Zaire 

E 

J. Burrows 2822 

T 

100 Acrostichum L. 
A. aureum L. 

45 

S 

Jaklitsch s.n. 

T 

HYMENOPHYLLACEAE 


106 Hymenophyllum Sm. 

H. capense Schrad. 

345 

E 

J. Burrows 2681 

LE 

H. polyanthos var. 
mossambicense sensu 
Schelpe 

H. capillare Desv. 

345 

E 

Chase 3187 

LE 

H. hirsutum sensu Schelpe 
& Diniz 

4 

E 

J. Burrows 2936 

L 

H. polyanthos Sw. 

var. kuhnii (C. Chr.) 
Schelpe 

34 

E 

Fisher & Schweickerdt 


H. henkelii Sim 
H. sibthorpioides (Bory ex 
Willd.) Mett. ex Kuhn 

34 

E 

321 

Muller 3248 

LE 

LE 

H. tunbridgense (L.) Sm. 

345 

E 

Chase 6678 

LE 


Check-list of pteridophytes of Zimbabwe 


201 


117 Trichomanes L. 


T. borbonicum v.d. Bosch 
T. erosum Willd. 

345 

E 

Chase 6573 

LE 

var. erosum 
var. aerugineum 

4 

E 

Chase 5851 

LE 

(v.d. Bosch) Schelpe 
T. pyxidiferum L. 

var. melanotrichum 

45 

E 

J. Burrows 2790 

L 

(Schlechtend.) Schelpe 

2345 

E 

Eyles 4451 

LE 

T. rigidum Sw. 

2345 

E 

Wild 3635 

T 


GLEICHENIACEAE 


144 Gleichenia Sm. 

G. polypodioides (L.) Sm. 345 NES Chase 6743 T 

G. umbraculifera (Kunze) 

Moore 5 E Wild 1475 T 

147 Dicranopteris Bernh. 

D. linearis (Burm.f.) 

Underw. 2345 NES Chase 7870 T 

POLYPODIACEAE 

162 Platycerium Desv. 

P. alcicome Desv. 4 

P. vassei Poisson 
P. bifurcatum sensu Sim 

163 Pyrrosia Mirb . 

P. lanceolata (L.) Farw. 4 

P. rhodesiana (C.Chr.) 

Schelpe 34 

P. schimperana (Mett. ex 
Kuhn) Alston 2345 

166 Microsorium Link 


Schelpe, E.A.C.L.E., in Con. FI. Angol., Pterid. Vol.: 121-124 (1977). 


M. pappei (Mett. ex Kuhn) 
Tardieu 

45 

E 

J. Burrows 2163 

LE 

M. punctatum (L.) Copel. 

345 

E 

Wild & Chase 5571 

LE 

M. scolopendrium (Burm.f.) 
Copel. 

45 

E 

Chase 6669 

TLE 

Phymatodes scolopendria 
(Burm.f.) Ching 

Pleopeltis Willd. 

P. excavata (Bory ex Willd.) 
Sledge 

2345 

NE 

Chase 6669 

LE 


E Biegel 2758 E 

E J. Burrows 2724 LE 

E Eyles 4472 LE 

NCES Chase 6577 LE 


202 


Journal of South African Botany 


Pleopeltis macrocarpa (Bory ex 


Willd.)Kaulf. 

2345 

CES 

Wild 4311 

LE 

P. schraderi (Matt.) Tardieu 

345 

ES 

Chase 6568 

LE 

186 Microgramma C-Presl 
M. lycopodioides (L.) 
Copel. 

245 

E 

J. Burrows 2725 

LE 

196 Belvisia Mirb. 

B. spicata (L.f.) Mirb. 

4 

E 

Chase 7196 

E 

206 Poly podium L. 

P. polypodioides (L.) 
Hitchcock 

subsp. ecklonii (Kunze) 
Schelpe 

345 

CES 

Wild 2119a 

LE 

GRAMMITIDACEAE 

210 Grammitis Sm . 

G. nanodes (A. Peter) Ching 4 E 

Mitchell 364 

LE 


211 Xiphopteris Kaulf. 

X. flabelliformis (Poir.) 


Schelpe 

345 

E 

Mitchell 366 

LE 

X. oosora (Bak.) Alston 

3 

E 

J. Burrows 2840 

L 

X. serrulata (Sw.) Kaulf. 

Congo 

E 

Williams 11 

L 

223 Loxogramme (Bl.) C.Presl. 

L. lanceolata (Sw.) C.Presl 

2345 

ES 

Chase 6665 

LE 


CYATHEACEAE 


228 CyatheaSm. 


Holttum, R. E. in Kew Bulletin 36 ( 3 ): 463-482 (1981) 
Tryon, R. in Contr. Gray Herb. 200 : 3-53 (1970) 


C. capensis (L.F.) Sm. 
Hemitelia capensis (L.f.) 
Kaulf. 

Alsophila capensis (L.f.) 
J.Sm. 

C. dregei Kunze 
Alsophila dregei (Kunze) 
Tryon 

C. manniana Hook. 
Alsophila manniana (Hook.) 
Tryon 

C. mossambicensis Bak. 
Alsophila mossambicensis 
(Bak.) Tryon 
C. thomsonii Bak. 

Alsophila thomsonii (Bak.) 
Tyron 


345 E J. Burrows 2547 

2345 NWCES Chase 6894 

234 E Chase 6696 

4 E J. Burrows 2788 

234 ES Chase 6864 


T 


T 

T 

T 

T 


Check-list of pteridophytes of Zimbabwe 


203 


DENNSTAEDTIACEAE 


239 Microlepia C.Presl 

M. speluncae (L.) Moore 12345 NWCES Mitchell 38 T 

242 Hypolepis Bernh. 

H. sparsisora (Schrad.) 


Kuhn 

345 

E 

Chase 6697 

244 Pteridium Scop. 

P. aquilinum (L.) Kuhn 
subsp. aquilinum 

2345 

NWCES Eyles 79 

subsp. centrali-africanum 
Hieron. 

234 

NWS 

Phipps 2881 

P. centrali-africanum 
(Hieron.) Alston 

245 Histiopteris (Agardh) J.Sm. 
H. incisa (Thunb.) J.Sm. 

235 

E 

Schelpe 5426 

248 Blotiella Try on 

B. glabra (Bory) Tyron 

245 

E 

Chase 6048 

B. natalensis (Hook.) Tryon 

2345 

ES 

Chase 3427 

252 Lindsaea Dryand. apud Sm. 
L. ensifolia Sw. 

45 

E 

J. Burrows 2864 

L. odorata Roxb. 

4 

E 

Ball 695 


THELYPTERIDACEAE 


Holttum, R. E. in J1 S. Afr. Bot. 40(2): 123-169 (1974) 

Schelpe, E.A.C.L.E., in Con. FI. Angol., Pterid. Vol: 149-160 (1977). 


rhelypteris Schmidel 
T. bergiana (Schlechtend.) 



Chase 6340 

LT 

Ching 

Amauropelta bergiana 
(Schlechtend.) Holtt. 

2345 

E 

T. chaseana Schelpe 
Christella chaseana 
(Schelpe) Holtt. 

T. confluens (Thunb.) 

235 

NCE 

Mitchell 584 

T 

Morton 

Dryopteris thelypteris sensu 
Sim 

T. palustris var. squamigera 
(Schlechtend.) Weath. 

2345 

NWCES Phipps 373 

T 


r H H H 


204 


Journal of South African Botany 


Thelypteris dentata (Forsk.) 
E. St John 
var. dentata 
Cyclosorus hispidulus 
A. Peter 

Christella dentata (Forsk.) 
Holtt. 

var. buchananii Schelpe 
T. friesii (Brause) Schelpe 
Christella friesii (Brause) 
Holtt. 

T. gueioziana (Mett.) 
Schelpe 

Christella gueinziana (Mett.) 
Holtt. 

T. hispidula (Decne.) Reed 
T. quadrangularis (Fee) 
Schelpe 

Christella hilsenbergii 
(C.Presl) Holtt. 

T. hilsenbergii (C.Presl) 
Schelpe 

Christella hispidula (Decne.) 
Holtt. 

T. interrupta (Willd.) 
K.Iwats, 

Cyclosorus interruptus 
(Willd.) H.Ito 
T. madagascariensis (Fee) 
Schelpe 

Pneumatopteris unita 
(Kunze) Holtt. 

T. oppositiformis (C.Chr.) 
Ching 

T. strigosa sensu Schelpe 
Amauropelta oppositiformis 
(C.Chr.) Holtt. 

T. pozoi (Lagasca) Morton 
Stenogramma pozoi 
(Lagasca) K.Iwats 
T. pulchra (Bory ex Willd.) 
Schelpe 

T. longicuspis (Bak.) 

Schelpe 

Pseudocyclosorus pulcher 
(Bory ex Willd.) Holtt. 

T. totta (Thunb.) Schelpe 
Dryopteris gongylodes 
(Schkuhr) Kuntze 
Cyclosorus tottus (Thunb.) 
P.-Sermolli 


NWCES Chase 6165 T 

Peter 30794 

E Mitchell 401 T 

E Chase 4628 T 

NWCES Chase 6167 T 

E Wild 2188 T 


NE Chase 3099 AT 

E Swynnerton 864 T 

E Schelpe 5670 LT 

E Chase 5667 LT 

CES Mitchell 745 T 


S J. Burrows 2657 AT 


2345 

45 

23 

345 

234 

123 

2345 

Tanzania 

35 

3 

345 


Check-list of pteridophytes of Zimbabwe 


205 


282 Ampelopteris Kunze 


A. prolifera (Retz.) Copel 2345 NWES 

ASPLENIACEAE 

288 Asplenium L. 

A. aethiopicum (Burm.f.) 

Chase 6540 

T 

Becherer 

A. praemorsum sensu Sim 
A. lividum sensu Schelpe 

2345 

NWCES Miller 2603 

TEL 

A. anisophyllum Kunze 

345 

E 

Chase 1011 

ELT 

A. atroviride Schelpe 

34 

E 

Schelpe 5446 

T 

A. auritum Sw. 

4 

E 

Chase 6558 

T 

A. blastophorum Hieron. 
A. boltonii Hook, ex 

2345 

E 

Swynnerton 845a 

T 

Schelpe 

2345 

ES 

Chase 2038 

ELT 

A. christii Hieron. 

5 

ES 

Fisher & Schweickerdt 
384 

T 

A. dregeanum Kunze 
A. erectum Bory ex Willd. 
var. usambarense 

2345 

E 

Chase 1115 

T 

(Hieron.) Schelpe 

2345 

ES 

Chase 6693 

T 

A. formosum Willd. 

234 

E 

Chase 3517 

ETL 

A. friesiorum C.Chr. 

2345 

E 

Mitchell 266 

ETL 

A. gemmiferum Schrad. 

345 

E 

Chase 6227 

TL 

A. holstii Hieron. 

34 

E 

Mitchell 377 

T 

A. hypomelas Kuhn 

Davallia hollandii Sim 
A. hollandii (Sim) C.Chr. 

34 

ES 

Fisher & Schweickerdt 
309 

Holland s.n. 

T 

A. inaequilaterale Willd. 

A. erectum Bory ex Willd. 
var. brachyotus (Kunze) 
Sim 

2345 

E 

Wild 2198 

LT 

A. linckii Kuhn 
A. lobatum Pappe & Raws. 

3 

E 

Chase 6576 

T 

var. lobatum 
A. erectum var. lobatum 
(Pappe & Raws.) Alston 
& Schelpe 

A. lunulatum var. gracile 
Sim 

var. pseudoabyssinicum 
Schelpe & 

345 

E 

Chase 3032 

LT 

N. C. Anthony 

4 

E 

Chase 7145 

ELT 

A. mannii Hook. 

234 

E 

Fisher & Schweickerdt 
219 

LE 

A. monanthes L. 

35 

E 

Chase 3047 

T 

A. mossambicense Schelpe 

4 

E 

Wild 2203 

T 

A. obscurum Bl. 

A. parablastophorum 

34 

E 

Mitchell 405 

TL 

A.Braithwaite 

0 

E 

Mitchell 381 

T 


206 


Journal of South African Botany 


Asplenium pellucidum Lam. 
subsp. pseudohorridum 


(Hieron.) Schelpe 

4 

E 

Mitchell 380 

LT 

A. preussii Hieron. 

A. pseudoaariciilatum 
Schelpe 

34 

E 

J. Burrows 1667 

T 

A. protensum Schrad. 

A. pumilum Sw. 

subsp. hymenophylloides 

2345 

E 

Chase 6690 

LT 

(Fee) Schelpe 
A. eylesii Sim 

23 

N 

Rodin 4360 
Eyles 564 

LT 

A. ramlowii Hieron. 

A. rutifolium (Berg.) Kunze 

24 

NWCES 

Chase 6316 

LT 

var. rutifolium 

var. bipinnatum (Forsk.) 

5 

E 

Ball s.n. 

LET 

Schelpe 

345 

ES 

Mitchell 282 

LET 

A. sandersonii Hook. 

345 

E 

Fisher 1098 

LE 

A. schelpei Braithwaite 
A. simii Braithwaite & 

5 

E 

Chase 5688 

LT 

Schelpe 

A. cuneatum. var. 

angustatum Sim 
A. theciferum (Kunth) 

Mett. 

var. concinnum (Schrad.) 

45 

E 

Chase 6274 

LE 

Schelpe 

2345 

ES 

Chase 2046 

LE 

A. trichomanes L. 

5 

E 

J. Burrows 2088 

TL 

A. unilaterale Lam. 

A. varians Wall, ex Hook. 
& Grev. 

subsp. fimbriatum 

3 

E 

Taylor 3380 

TL 

(Kunze) Schelpe 
291 Ceterach DC. 

45 

ES 

Chase 5124 

LT 

C. cordatum (Thunb.) Desv. 
Asplenium cordatum 
(Thunb.) Sw. 

304 Athyrium Roth. 

A. scandicinum (Willd.) 
C.Presl 

15 

NWCES 

Miller 5501 

T 

var. scandicinum 

345 

E 

Chase 2199 

T 

var. rhodesianum Schelpe 

0 

E 

Schelpe 5722 

T 

A. schimperi Moug. ex Fee 
307 Diplazium Sw. 

235 

CE 

Chase 3774 

T 

D. nemorale (Bak.) Schelpe 
D. zanzibaricum (Bak.) 

4 

E 

Chase 5700 

T 

C.Chr. 

321 Dryoathyrium Ching 

245 

E 

Chase 6181 

T 

D. boryanum (Willd.) Ching 
Athyrium glabratum (Mett. 

4 

E 

Chase 6182 

T 


ex Kuhn) Alston 
Parathyrium boryanum (Willd.) 


Holtt. 


Check-list of pteridophytes of Zimbabwe 207 


328 Ctenitis (C.Chr.) C.Chr. ex Tardieu 




C. cirrhosa (Schum.) Ching 

234 

E 

Chase 5127 

T 

C. lanuginosa (Willd. ex 
Kaulf.) Copel. 

345 

E 

J. Burrows 2122 

T 

Dryopteris lanuginosa 
(Willd. ex Kaulf.) C.Chr. 

337 Tectaria Cav. 

T. gemmifera (Fee) Alston 

2345 

E 

Chase 5715 

T 

353 Didymochlaena Desv. 

D. truncatula (Sw.) J.Sm. 

345 

E 

Gilliland 1747 

T 

356 Polystichum Roth. 

P. luctuosum (Kunze) 
Moore 

5 

E 

Chase 1083 

T 

P. transvaalense 
N.C. Anthony 

35 

E 

Chase 7482 

T 

P. setiferum sensu Schelpe 
P. zambesiacum Schelpe 

34 

E 

Schelpe 5751 

T 

364 Arachniodes Bl. 

A. foliosa (C.Chr.) Schelpe 

25 

E 

J. Burrows 2227 

T 

Dryopteris foliosa C.Chr. 

369 Dryopteris Adans. 

D. athamantica (Kunze) 
Kuntze 

2345 

CES 

Mitchell 45 

T 

D. callolepis C.Chr. 

5 

E 

Chase 5262 

T 

D. inaequalis (Schlechtend.) 
Kuntze 

2345 

NCES 

Chase 6531 

T 

D. kilemensis (Kuhn) 
Kuntze 

234 

E 

Chase 5390 

T 

D. lastii (Bak.) C.Chr. 
D. manniana (Hook.) 
C.Chr. 

34 

E 

J. Burrows 1676 

T 

D. squamiseta (Hook.) 
Kuntze 

25 

E 

Chase 4640 

T 

D. buchananii (Bak.) 
Kuntze 

D. wallichiana (Spreng.) 
Hylander 

Malagasy 

E 

Chase 7337 

T 

374 Bolbitis Schott 

B. gemmifera (Hieron.) 
C.Chr. 

Rep. 

Angola 

E 

Biegel 2827 

T 

B. heudelotii (Bory ex Fee) 
Alston 

245 

NE 

Chase 6638 

LA 


208 


Journal of South African Botany 


380 Lomariopsis Fee 

L. warneckei (Hieron.) 

Alston 34 E J. Burrows 2116 E 

L. nigrescens Holtt. 

381 Elaphoglossum Schott 


E. acrostichoides (Hook. & 


Grev.) Schelpe 
E. conforme var. lineatum 
(Kuhn ex Christ) C.Chr. 

345 

E 

Chase 3792 

EL 

E. aubertii (Desv.) Moore 

345 

E 

Chase 4509 

L 

E. chevalieri Christ 

34 

E 

Chase 6667 

EL 

E. deckenii (Kuhn) C.Chr. 

3 

E 

J. Burrows 2841 

EL 

E. hybridum (Bory) Brack. 

345 

E 

Chase 6666 

ETL 

E. kuhnii Hieron. 

3 

CE 

Schelpe 5647 

TL 

E. lastii (Bak.) C.Chr. 
E. macropodium (Fee) 

34 

E 

Fisher & Schweickerdt 
326 

E 

Moore 

E. conforme var. latifolium 
sensu Sim 

345 

E 

Schelpe 5742 

LE 

E. marojejyense Tardieu 

E. petiolatum (Sw.) Urb. 
subsp. salicifolium (Willd. 

Malagasy 

Rep. 

E 

Mitchell 367 

L 

ex Kaulf.) Schelpe 
E. spathulatum (Bory) 

234 

E 

Chase 3308 

EL 

Moore 

E. spathulatum var. 

2345 

NE 

Drummond 6049 

L 


uluguruense (Reim.) 

Schelpe 

DAVALLIACEAE 

388 Davallia Sm. 

D. chaerophylloides (Poir.) 



Steud. 

45 

E 

Chase 2190 

EL 

394 

Rumohra Raddi 
R. adiantiformis (Forst.) 
Ching 

5 

E 

J. Burrows 2262 

T 

395 

Oleandra Cav. 

O. distenta Kunze 

2345 

NCE 

Chase 6912 

ETL 

396 

Arthropteris J.Sm. 

A. monocarpa (Cordem.) 
C.Chr. 

2345 

CE 

Chase 4499 

ETL 


A. orientalis (J.F.Gmel.) 
Posth. 

234 

NWCES 

Chase 6406 

TL 

398 

Nephrolepis Schott 
N. biserrata (Sw.) Schott 

245 

E 

Biegel 2779 

ETL 


N. undulata (Afz. ex Sw.) 
J.Sm. 

234 

NWCE 

Davies 978 

T 


Check-list of pteridophytes of Zimbabwe 209 

BLECHNACEAE 


399 Blechnum L. 


B. attenuatum (Sw.) Mett. 

235 

E 

Chase 6645 

EL 

B. australe L. 

5 

E 

Fisher 1257 

T 

B. giganteum (Kaulf.) 





Schlechtend. 

345 

CE 

Swynnerton 835a 

LT 

B. attenuatum 





var. giganteum 





B . inflexum (Kunze) Kuhn 

5 

E 

Fisher 1428 

T 

B. ivohibense C.Chr. 

4 

E 

J. Burrows 2839 

T 

B. punctulatum Sw. 

35 

NS 

J. Burrows 2695 

TL 

B. sylvaticum Schelpe 

345 

E 

Chase 4268 

T 

B capense (L.) Schlechtend. 





B. tabulare (Thunb.) Kuhn 

2345 

CE 

Chase 7134 

T 

409 Stenochlaena J.Sm. 





S. tenuifolia (Desv.) Moore 

45 

E 

J. Burrows 2780 

TC 

MARSILEACEAE 



410 Marsiiea L. 





M. apposita Launert 

15 

NW 

Gibbs 289 

A 

M. coromandelina Willd. 

5 

S 

Ngoni 137 

A 

M. ephippiocarpa Alston 

125 

NCES 

Rendle 307 

A 

M. farinosa Launert 

15 

wc 

Plowes 1878 

A 

M. macrocarpa C.Presl 

5 

w 

Whellan 1664 

A 

M. minuta L. 

124 

N 

J. Burrows 2521 

A 

M. nubica A. Braun 

15 

NW 

Gibbs-Russell 1629 

A 

SALVINIACEAE 



413 Salvinia Adans. 





* S. molesta Mitchell 

1245 

NWCE 

Mitchell 1330 

A 

S. auriculata sensu Schelpe 






AZOLLACEAE 



414 AzollaLam. 





A. filiculoides Lam. 

5 

NC 

J. Burrows 3066 

A 

A. nilotica Decne. ex Mett. 

234 

NW 

Guy 902 

A 

A. pinnata R.Br. 

1245 

W 

Van Someren 87308 

A 


var. africana (Desv.) 
Bak. 


References 

Anthony, N. C. and Schelpe, E. A., 1982. New species and new combinations in 
African Pteridophyta and Orchidaceae. Contr. Bol. Herb. 10. 

Braithwaite, A. F., 1972. Two new species of Asplenium from Southern Africa. JJ 
S. Afr. Bot. 38(1): 1-7. 

Crabbe, J. A., Jermy, A. C. and Mickel, J. T„ 1975. A new generic sequence for 
the pteridophyte herbarium. Fern Gaz. 11 (2 & 3): 141 162. 


210 


Journal of South African Botany 


Holttum, R. E., 1974. Thelypteridaceae of Africa and adjacent islands. Jl S. Afr. 
Bot. 40(2): 123-169. 

Holttum. R. E., 1981. The tree-ferns of Africa. Kew Bull. 36(3): 463-482. 

Kornas, J., 1979. Distribution and ecology of the pteridophytes in Zambia. Warsaw: 
Panstwowe Wydawnictwo Naukowe. 

Schelpe, E. A., 1969. A revised check-list of the Pteridophyta of Southern Africa. Jl 
S. Afr. Bot. 35(2): 127-140. 

Schelpe, E. A., 1969. Reviews of Tropical African Pteridophyta 1. Contr. Bol. Herb. 

10 . 

Schelpe, E. A., 1970. Pteridophyta. In: Flora Zambesiaca. London: Crown Agents. 
Schelpe, E. A., 1977. Pteridophyta. In: Conspectus Florae Angolensis. Lisbon. 
Schelpe, E. A., 1979. A new name for Blechnum capense (L.) Schlechtend. Jl S. 
Afr. Bot. 45(2): 221. 

Schelpe, E. A. and Diniz, M. A., 1979. Pteridophyta. In: Flora de Mozambique. 
Lisbon. 

Tryon, R., 1970. The classification of the Cyatheaceae. Contr. Gray Herb. No. CC. 


INDEX 


Genus Page 

No. 

100 Acrostichum 200 

039 Actiniopteris 197 

ADIANTACEAE 197 

086 Adiantum 199 

228 Alsophila 202 

260 Amauropelta 203, 204 

282 Ampelopteris 205 

035 Anemia 197 

062 Anogramma 199 

364 Arachniodes 207 

396 Arthropteris 208 

050 Aspidotis 199 

ASPEENIACEAE 205 

288 Asplenium 205,206 

304 Athyrium 206 

414 Azolia 209 

AZOLLACEAE 209 

196 Belvisia 202 

BLECHNACEAE 209 

399 Blechnum 209 

248 Blotiella 203 

374 Bolbitis 207 

037 Ceratopteris 197 

291 Ceterach 206 


Page 


047 Cheilanthes 198 

260 Christella 203,204 

328 Ctenitis 207 

228 Cyathea 202 

CYATHEACEAE 202 

260 Cyclosorus 204 

288 Davallia 205 

388 Davallia 208 

DAVALLIACEAE 208 

DENNSTAEDTIACEAE 203 

147 Dicranopteris 201 

353 Didymochlaena . . . . .... 207 

307 Dipiazium 206 

047 Doryopteris 198 

321 Dryoathyrium . . . . 206 

260 Dryopteris 203,204 

328 Dryopteris 207 

364 Dryopteris 207 

369 Dryopteris 207 

381 Elaphoglossum 208 

EQUISETACEAE 196 

011 Equisetum 196 

144 Gleichenia 201 


Journal of South African Botany 211 


Page 


GLEICHENIACEAE 201 

GRAMMITIDACEAE 202 

210 Grammitis 202 

063 Gymnogramma 199 

228 Hemitelia 202 

245 Histiopteris 203 

HYMENOPHYLLACEAE 200 
106 Hymenophyllum 200 

050 Hypolepis 199 

242 Hypolepis 203 

ISOETACEAE 196 

009 Isoetes 196 

252 Lindsaea 203 

380 Lomariopsis 208 

223 Loxogramme 202 

LYCOPODIACEAE 195 

003 Lycopodium 195 

034 Lygodium 197 

022 Marattia 197 

MARATTIACEAE 197 

410 Marsilea 209 

MARSILEACEAE 209 

186 Microgramma 202 

239 Microlepia 203 

166 Microsorium 201 

036 Mohria 197 

398 Nephrolepis 208 

047 Notholaena 198 

395 Oleandra 208 

OPHIOGLOSSACEAE 196 

016 Ophioglossum 196 

026 Osmunda 197 

OSMUNDACEAE 197 


197 


Page 


321 Parathyrium 206 

047 Pellaea 198 

057 Pellaea 199 

166 Phymatodes 201 

063 Pityrogramma 199 

162 Platycerium 201 

185 Pleopeltis 201,202 

260 Pneumatopteris 204 

POLYPODIACEAE 201 

206 Polypodium 202 

356 Polystichum 207 

260 Pseudocyclosorus 204 

PSILOTACEAE 195 

001 Psilotum 195 

244 Pteridium 203 

096 Pteris 200 

163 Pyrrosia 201 

394 Rumohra 208 

413 Salvinia 209 

SALVINIACEAE 209 

032 Schizaea 197 

SCHIZAEACEAE 197 

008 Selaginella 196 

SELAGINELLACEAE 196 

409 Stenochlaena 209 

260 Stenogramma 204 

337 Tectaria 207 

THELYPTERIDACEAE 203 

260 Thelypteris 203 

029 Todea 197 

117 Trichomanes 201 


093 Vittaria . . . 
211 Xiphopteris 


PARKERIACEAE 




J1 S. Afr. Bot. 49 (3): 213-219 (1983) 


A REMARKABLE NEW LEUCOSPERMUM (PROTEACEAE) FROM THE 
SOUTHERN CAPE 

J. P. Rourke 

(Compton Herbarium , Kirstenbosch, Private Bag X7, Claremont 7735, 
R.S.A.) 

Abstract 

Leucospermum hamatum Rourke (Proteaceae), a new species from the southern 
Cape is described, characterised by its minute 4 to 7-flowered inflorescences and re- 
trorsely barbed subterminalregion of the style. 

UlTTREKSEL 

'N BESONDERE NUWE LEUCOSPERMUM (PROTEACEAE) VANAF DIE 
SUID-KAAP 

Leucospermum hamatum Rourke (Proteaceae), ’n nuwe soort vanaf die Suid-Kaap 
word beskryf, gekenmerk deur die klein 4-7 blomme in die bloeiwyses en na onder- 
gerigte weerhake aan die subterminale streek van die styl. 

Key words: Leucospermum , sp. nov., Proteaceae, southern Cape. 

This extraordinary species was discovered in 1978 on the north slopes of 
the Outeniqua mountains in the southern Cape, by Mr. G. le Roux, forester 
at Ruitersbos Forest Station. He brought it to the attention of Mr. W. J. 
Bond of the Saasveld Forest Research Station near George, who in turn 
submitted herbarium material to me for identification. 

As Bond's original collection consisted of largely sterile, vegetative ma- 
terial, it was determined as Leucospermum secundifolium Rourke, since the 
foliage closely resembled that of the latter species. Subsequently however, 
Mr. J. H. J. Vlok, also of the Saasveld Forest Research Station, queried this 
identification after obtaining flowering specimens. He thereupon invited me 
to accompany him on a field excursion to examine the species in situ during 
the height of the flowering season, in August 1982. After examining a popu- 
lation in the field, it was at once apparent that this bizarre species was quite 
unlike Leucospermum secundifolium, and, indeed, bore little resemblance 
to any other known Leucospermum (Rourke, 1972). It is here described as 
new. 

Leucospermum hamatum Rourke, species nova; species nana miranda, a 
habitu prostrato implexo, foliis secundis, inflorescentiis minutis 4-7 floribus, 


Accepted for publication 20th January, 1983. 


214 


Journal of South African Botany 


perianthiis curtis proximalibus inflatis, stylis hamatis retrorsis distalibus et 
valde adaxialibus arcuatis, distinguitur. 

Frutex humilis prostratis; tegetes densae formantes, 1-3 m in diam.; ad 
100 mm altae. Folia secunda, sparsim puberula, postea glabra; anguste ob- 
longa vel oblonga-elliptica, 45-65 mm longa, 8-18 mm lata; breviter petio- 
lata, petioli ad 5 mm longi; apices plerumque tridentati, aliquando integri. 
Inflorescentia turbinata, 15-20 mm in diam., 4-7 floribus. Involucrum redac- 
tum ad 3 vel 4 bracteas. Bracteae ovatae acuminatae, conduplicatae, 10- 
15 mm longae, 4-6 mm latae, sparsae puberulae. Perianthium 10-14 mm 
longum, valde adaxile arcuatum, utriculosum et inflatum distale, glabrum 
sed unguibus marginibus et sinibus fimbriatis. Stylus 18-20 mm longus, cyg- 
neus, valde adaxilis arcuatus, barbatus retrorsus apicem versus. Stigma coni- 
ca acuta, 20 mm longa. 

Typos: Cape Province — 3322 (Oudtshoorn): Outeniqua mountains, north slopes of 
Moordkuils River catchment area on Klein Moeras River Farm (-CC), 24/08/1982, J. 
P. Rourke 1779 (NBG, holotypus; PRE, K, MO, S, STE, BOL, L, NSW, SAAS, 
PRF, isotypus). 


A prostrate sprawling shrub forming dense mats 1-3 m in diam., up to 
100 mm in height with several stout (to 25 mm thick) branches radiating 
horizontally along ground, from a single stout main trunk up to 45 mm in 
diam. Flowering branches slender, trailing, frequently dichotomous, 2 mm in 
diam., very sparsely villous, usually bearing a single terminal inflorescence 
and several axillary inflorescences. Leaves secund, narrowly oblong to ob- 
long-elliptic, 45-65 mm long, 8-18 mm wide, shortly petiolate, petiole to 
5 mm long; apex usually tridentate occasionally entire; sparsely puberulous, 
later glabrous. Inflorescence turbinate, 15-20 mm in diam., consisting of 4 to 
7 flowers apparently arranged in a single whorl; pedunculate, peduncle 
10-20 mm long. Involucre either absent or reduced to 3 or 4 bracts, lanceo- 
late-acute to acuminate, 7-10 mm long, 2-8 mm wide, sparsely puberulous. 
Floral bracts ovate acuminate, conduplicate, tightly clasping the base of the 
perianth, 10-15 mm long, 4-6 mm wide, sparsely puberulous externally, 
margins puberulously fimbriate. Perianth very strongly adaxially curved, 
10-14 mm long; tube utriculose, much inflated and swollen distally to 5 mm 
in diam., glabrous; claws very reduced and tightly adaxially curved, broad- 


Fig. 1. 

Leucospermum hamatum, drawn from the type material (Rourke 1779). 

(1) flowering shoot; (2) single flower and subtending floral bract; (3) perianth spread 
open; (4) gynoecium showing style, pollen presenter and hypogynous scales; (5) 
floral bract; (6) perianth limb and anther; (7) fruit, adaxial view. 


New Leucospermum (Proteaceae) from southern Cape 


215 





216 


Journal of South African Botany 


ened proximally, tapering distally, the three adaxial perianth segments in- 
completely fused, 4 mm long, the free abaxial segment 5-6 mm long; outer 
surface glabrous but margins and the two adaxial sinuses at junction with 
perianth tube lanately fimbriate; limbs lanceolate-cymbiform, acute, 2 mm 
long, outer surface granulate-glabrescent. Anthers 4, sessile, arising from a 
prominently swollen cushion of carnose connective. Style 18-20 mm long, 
strongly adaxially cygneous, subterminal region tapered, retrorsely barbed, 
becoming sparsely granulate basally. Pollen presenter conic-acute, 2 mm 
long, stigmatic groove terminal. Ovary not differentiated from style, min- 
utely puberulous 1 mm long. Hypogynous scales linear-obtuse, 2 mm long. 
Fruit a minutely puberulous, cylindric, greyish-white achene, broadly emar- 
ginate at base. 

Relationships and Diagnostic Characters 

This species is not obviously allied to any other in the genus, nor can it be 
accommodated in any existing section within Leucospermum (Rourke, 
1972). The inflorescence is so reduced as to consist of between four and 
seven (usually five) flowers, arranged in a radially symmetric manner so as 
to form an apparent whorl, with the styles of each flower all strongly curved 
inwards. Leucospermum hamatum is probably both the most reduced and 
the most specialised species in the genus. In each inflorescence the involucre 
is either absent or reduced to a mere three or four bracts, while the floral 
bracts subtending each flower are enlarged to form a cup-like “pseudo-in- 
volucre’' clasping the perianths. 

The specific epithet “hamatum” draws attention to a character unique in 
Leucospermum and in fact in the whole of the South African Proteaceae; 
namely, the presence of minute, retrorsely directed barbs in the subterminal 
region of the style. 

Distribution, Habitat and Biology 

Leucospermum hamatum is apparently a highly localised endemic. It is 
presently known from a single site in the catchment area of the Moordkuils 
River on Klein Moeras River farm in the northern foothills of the Outeni- 
qua mountains. When inspected in August 1982, a population of well over a 
thousand individuals was found at this locality, dispersed over several hec- 
tares at a mean elevation of 750 m. The plants were in various stages of de- 
velopment; some being fully mature, while others were seedlings regenerat- 
ing after a recent fire. Leucospermum hamatum is a seed-regenerator which 
does not survive burning. 

Mature shrubs develop into dense, prostrate, sprawling mats up to 3 m in 
diameter with numerous horizontally radiating branches arising from a 
single stout main trunk. Most specimens attain less than 100 mm in height 


New Leucospermum (Proteaceae) from southern Cape 


217 



Fig. 2. 

Distribution of Leucospermum hamatum. 



Fig. 3. 

Growth habit showing the prostrate mode of growth. 


DDD1 


218 


Journal of South African Botany 


when fully grown and consequently tend to be rather inconspicuous in the 
rank surrounding fynbos, where tall, shrubby associates, such as Leucaden- 
dron ericaefolium , Berzelia intermedia and Erica seriphiifolia are prominent. 

This species was observed in a variety of niches, such as on moist clay 
derived from the lower Table Mountain Sandstone shaleband, in damp 
peaty seepage zones, as well as on drier, rocky, well-drained north-facing 
Table Mountain Sandstone slopes. Normally the foliage is bright green but 
leaves on new shoots and on those exposed to direct sunlight, tend to be suf- 
fused with deep carmine tones. 

Flowering takes place over several months, principally between July and 
November, but as the tiny (15-20 mm in diam.) inflorescences are borne 



Fig. 4. 

Developing inflorescences on a flowering shoot. The unopened perianths and elon- 
gating styles are clasped within an “involucre” formed by the subtending floral bracts. 

X 2. 

at ground level, often concealed in the dense surrounding fynbos, they are 
easily overlooked. The inconspicuousness of the inflorescences is heightened 
by their drab colouration; the perianths are hyaline, the styles pale yellow- 
amber and the “involucre” of floral bracts is dull carmine. It thus seems un- 
likely that this species is bird-pollinated or even insect-pollinated, as are the 
majority of leucospermums (Rourke, 1972). 

Pollination by small rodents is, however, a distinct possibility although 
this supposition is presently based on circumstantial evidence. Nectar wells 


New Leucospermum (Proteaceae) from southern Cape 


219 


up and accumulates in the swollen, bladder-like perianth tubes. In many 
inflorescences the walls of these inflated perianth tubes were observed to 
have been gnawed away, presumably to provide access to the nectar reser- 
voirs. A small rodent placing its snout into the centre of each inflorescence 
would be perfectly positioned so as to gnaw through the inflated perianth 
tubes and lap the nectar while brushing pollen from the inwardly directed 
pollen presenters onto its snout. 

In almost a third of the recently opened inflorescences, which were 
examined in the field, the inflated perianth tubes of most of the flowers had 
been punctured, gnawed (by rodents ?), and the nectar removed in this 
manner. 

Specimens Examined 

CAPE PROVINCE — 3322 (Oudtshoorn): Outeniquas, Klein Moeras River catch- 
ment at foot of steep north slope (-CC), 29/5/1978, W. J. Bond 1414 (NBG, SAAS); 
Outeniqua mountains, north slopes of Moordkuils River catchment area on Klein 
Moeras River farm, 24/8/1982, J. P. Rourke 1779 (NBG, PRE, K, MO, S, STE, 
BOL, L, NSW, SAAS, PRF). 

Acknowledgements 

I would like to thank Mr. J. H. J. Vlok and Mr. G. le Roux of the De- 
partment of Environment Affairs (Directorate of Forestry) for arranging the 
field excursion w'hich enabled me to collect and study Leucospermum hama- 
tum in its natural habitat. Their enthusiasm and assistance in undertaking 
this work is gratefully acknowledged. 

References 

Rourke, J. P., 1972. Taxonomic studies on Leucospermum R.Br. .// S. Afr. Bot. 
Suppl. Vol. 8: 1-194. 




J1 S. Afr. Bot. 49 (3): 221-242 (1983) 


THE TAXONOMIC VALUE OF TRICHOMES IN 
PELARGONIUM L’HERIT. (GERANIACEAE) 


L. OOSTHUIZEN* 

( Department of Botany, University of Stellenbosch, Stellenbosch 7600, 
R.S.A.) 


Abstract 

A detailed study of the indumentum of the leaves of 133 Pelargonium species led 
to the classification of the trichomes into 11 different types: six non-glandular and 
five glandular hair types. The non-glandular hairs are uni- to multicellular and of dif- 
ferent lengths, shape and wall thickness. Some hairs have a podium at the base. The 
glandular hairs consist of a uniserial stalk of various lengths and unicellular head of 
various shapes. The distribution of the various trichome and indumentum types in 
the different taxa of Pelargonium is discussed, especially where it can be of diagnos- 
tic value and serve as possible indication of relationships. 


UlTTREKSEL 

DIE TAKSONOMIESE WAARDE VAN TRIGOME IN PELARGONJUM 
L'HERIT. (GERANIACEAE) 

'n Volledige ondersoek van die indumentum van die blare van 133 Pelargonium- 
spesies het gelei tot die klassifikasie van die trigome in die genus in 11 tipes: ses 
haartipes en vyf klierhaartipes. Die hare is een- tot veelsellig en varieer in lengte, 
vorm en wanddikte. Sommige hare het ’n podium by die basis. Die klierhare bestaan 
uit ’n uniseriale steel van verskillende lengtes en eensellige kop van verskillende 
vorme. Die verspreiding van die verskillende trigoom- en indumentumtipes in die 
verskillende taksa van Pelargonium word bespreek, veral waar dit van diagnostiese 
waarde kan wees en as moontlike aanduiding van verwantskappe kan dien. 


Key words: Pelargonium , Geraniaceae, trichomes, diagnostic characters. 
Introduction 

Trichomes can be studied from two different perspectives: (1) the nature 
of the individual trichomes themselves, and (2) the characteristics which 
they collectively impart to the surfaces upon which they occur, i.e. the 
nature of the indumentum layer as a whole. Since the indumentum is not as 


Present address: Department of Botany, University of Pretoria, l letoria 000 
R.S.A. 

Accepted for publication 24th January, 1983. 

221 


222 


Journal of South African Botany 


much influenced by the structure of the individual trichomes' as by their dis- 
tribution and density, trichomes of rather different structure may give rise to 
similar indumentum types. It is also possible that the environment has a 
greater influence in modifying the indumentum than in changing the type of 
trichome (Johnson, 1975). 

According to Carlquist (1961) trichomes often provide the taxonomist 
with the most important anatomical characters used in systematic compari- 
sons. This is probably due to their variety, almost universal presence in 
angiosperms, simple preparation methods for studying, and close resem- 
blance to patterns of variation in classification systems. In a taxonomic tri- 
chome study it is necessary to examine the whole trichome complement, and 
indicate the organographic distribution of the different trichome types as 
well. 

The last comprehensive taxonomic revision of the genus Pelargonium 
was done by Knuth in 1912. It has become clear that his identification keys 
are unsatisfactory, and that some species are erroneously placed into sec- 
tions. Since his revision, a number of new species have also been described. 
The genus includes more than 200 natural species of which the majority oc- 
cur in southern Africa. 

Phillips (1951), Willis (1973) and Van der Walt (1977) mention the fact 
that Pelargonium species are often conspicuously hairy and aromatic, and 
produce essential oils used in the perfume industry. Because trichome 
studies have already proved to be of taxonomic value in a number of taxa of 
various ranks, this study has been carried out as part of the extensive taxo- 
nomic project on Pelargonium in progress at the University of Stellenbosch. 

The purpose of this study was to distinguish between, and classify, differ- 
ent trichome and indumentum types present in the genus, and to determine 
whether these types can be used as taxonomic characters in distinguishing 
between taxa, and in the determination of the possible phylogenetic re- 
lationships of these taxa. 


Material and Methods 

The petioles and laminae of 133 species and subspecies of Pelargonium 
were studied. The material was obtained from plants cultivated in the bota- 
nic garden of the University of Stellenbosch. In order to describe the type 
and density of the indumentum, the petioles and laminae were studied with 
the aid of a dissection microscope. Transverse sections were made of the 
above-mentioned organs in order to examine the structure of the individual 
trichomes. The figures presented in this paper are semi-schematic and drawn 
to scale. 


Taxonomic value of trichomes in Pelargonium 


223 


Indumentum types 

The following terminology has been used to describe the indumentum 
types in the different taxa of Pelargonium : 
ciliated: thin hairs, all of approximately the same length, on the margin of 
the leaf 

fimbriate: long, relatively thick, filamentous hairs on the margin of the 
leaf 

glabrous: without trichomes 
glandular: covered with glandular hairs 

hirsute: covered with long, straight, stiff hairs (not spiny) (hairs are thin- 
ner than in the case of a hispid indumentum) 
hirtellous (minutely hirsute): like hirsute, but with shorter hairs 
hispid: covered with long, straight, stiff, almost spiny hairs (hairs are 
thicker than in the case of a hirsute indumentum) 
lanate (woolly): covered with long, dense, entangled curly hairs - longer 
than in the case of a tomentose indumentum, and without a matted 
appearance 

pilose: covered with thin, soft hairs, longer than in the case of a pubescent 
indumentum 

pubescent: covered with short, thin, soft hairs 

setose (bristly): covered with very stiff, hard hairs (hairs are thicker and 
harder than in the case of a hispid indumentum) 
strigose: covered with stiff hairs with a sharp point, the hairs being ap- 
pressed to the surface and orientated in a distal or proximal direction 
tomentose (densely woolly): covered with very dense, entangled, short 
hairs with a matted appearance 

velutinous: covered with very dense, straight, short hairs - appears vel- 
vety 

villous: covered with long, thin, soft hairs - not as dense as a tomentose 
indumentum, and without a matted appearance. 


Density of indumentum 

The following terms were used to describe the density of the indumentum 
(given in descending degree of density): 

dense (to very dense): epidermal cells are hardly visible 
relatively dense: trichomes are abundant 
scattered: trichomes are less abundant 
sparse (to almost glabrous): trichomes are scarce. 

For each of the examined taxa the indumentum type and density were de- 
scribed in detail for the petiole, and ad- and abaxial sides of the lamina. 


224 


Journal of South African Botany 



Taxonomic value of trichomes in Pelargonium 


225 


Results 
Trichome types 

Only uni- and multicellular non-glandular hairs, and glandular hairs with a 
uniserial stalk (of various lengths) and unicellular head (of various shapes) 
occur in Pelargonium species. The length, thickness of the wall, shape 
(straight or curly) and presence of a podium served as criteria to classify the 
non-glandular hairs. (A podium is a multicellular structure, consisting of 
epidermal cells which are raised above the rest of the surface, and which 
surround the hair base, which is sunken into the podium.) With glandular 
hairs, the length of the stalk and shape of the glandular head and basal cell 
were used as characters of distinction. Due to the fact that transitional forms 
also occur, it is often difficult to classify such hairs. Glandular hairs, which 
are more closely connected to physiological processes in the plant, exhibit 
greater constancy of characters. 

The following trichome types occur in the different taxa of Pelargonium 
(Fig. 1): 

Type 1: short, straight, stiff hair; (a) without a podium; (b) with a podi- 
um. 

Type 2: long, straight, stiff to almost spiny hair; (a) without a podium; (b) 
with a podium. 

Type 3: very stiff, hard hair with a prominent podium. 

Type 4: short, thin, soft hair. 

Type 5: long, thin, soft hair; (a) without a podium; (b) with a podium. 

Type 6: curly hair. 

Type 7: short glandular hair with a small globular head, and stalk consist- 
ing of two to four cells. 

Type 8; short glandular hair with a large globular head, and stalk consist- 
ing of one to four cells. 

Type 9: short glandular hair with a small bulb-shaped head, and stalk con- 
sisting of two to five cells. 

Type 10: relatively long to very long glandular hair with a globular or 


Fic. 1. 

Trichome types occurring in Pelargonium. 1: Short, straight, stiff hair (a) without a 
podium, (b) with a podium. 2: Long, straight, stiff to spiny hair (a) without a 
podium, (b) with a podium. 3: Very stiff, hard hair with a prominent podium. 4. 
Short, thin, soft hair. 5: Long, thin, soft hair (a) without a podium, (b) with a po- 
dium. 6: Curly hair. 7: Short glandular hair with a small globular head and stalk 
consisting of two to four cells. 8: Short glandular hair with a large globular hea > an 
stalk consisting of one to four cells. 9: Short glandular hair with a small bulb-s ape 
head and stalk consisting of two to five cells. 10: Relatively long to very rig g ancu 
lar hair without a globular or bulb-shaped head, and stalk consisting of three to nu 
merous cells of which the basal one is always elongated and often swollen. . or 
glandular hair with a large pear-shaped head. 


226 


Journal of South African Botany 


bulb-shaped head, and stalk consisting of three to numerous cells of 
which the basal one is always elongated and often swollen. 

Type 11: short glandular hair with a large pear-shaped head. 

A total of 11 trichome types have thus been distinguished: types 1 to 6 are 
non-glandular hair types, and 7 to 11 glandular hair types. Types 1. 2 and 4 
have been subdivided on the ground of the presence of a podium surround- 
ing the hair base. In many cases the sculpture of the hair surface also varies; 
a scanning electron microscopic study may therefore lead to a more detailed 
classification of trichome types. 

Table 1 represents the occurrence of the different trichome types on the 
lamina and petiole of 133 taxa of Pelargonium. In the columns the trichome 
types are represented by their numbers (1 to 11) as indicated in Fig. 1. 
Even - row contains the trichome types of a single species. The table is pre- 
sented as illustration of how taxonomic conclusions could be drawn from 
this trichome study. 

Page 

Table 227-240 


Taxonomic value of trichomes in Pelargonium 


227 


Table 1 

Occurrence of trichomes on the petiole and lamina of species of Pelargonium. 
Trichome numbers refer to the types as indicated in Fig. 1. Unidentified species have 
been given a project number, and are referred to as P. species (project number x). 


Species 

Petiole 

Non-Glandular Hairs 

Glandular Hairs 


la 

lb 

2a 

2b 

3 

4 

5a 

5b 

6 

7 

8 

9 

10 

11 

Section: 















! 

Hoarea 
















P. auritum 
















-subsp. auritum 

* 



* 












-subsp. carneum 

* 

* 


* 












P. carneum 

* 



* 

* 











P. incrassatum 




* 


* 








* 


P. oblongatum 






* 


* 






* 


P. pinnatum 

* 



* 












P. punctatum 

* 



• 












P. rapaceum 
















P. spathulatum 
















P. species (1571) 

x 

* 


• 












P. species (2587) 

* 

* 


* 












P. species (1613) 




* 












P. species (442) 
















P. species (2440) 

* 















P. species (1456) 




* 












P. species (1653) 

* 















P. species (2439) 

* 















P. species (2443) 

* 















P. species (1454) 

X 















Section 
















Sevmouria 
















P. asarifolium 
















P. dipetalum 















P. marginatum 















P. trifoliatum 














1 


228 


Journal of South African Botany 


Species 

Lamina 

Non-Glandular Hairs 

Glandular Hairs 


la 

lb 

2a 

2b 

3 

4 

5a 

5b 

6 

7 

8 

9 

10 

11 

Section: 















Hoarea 















P. auritum 















-subsp. auritum 




* 






* 

* 


* 


-subsp. carneum 




* 






* 

* 




P. carneum 




* 






* 





P. incrassatum 




* 


* 

* 



* 

* 


* 


P. oblongatum 






* 

* 



* 



* 


P. pinnatum 


* 


* 






• 





P. punctatum 

* 



* 






¥ 



* 


P. rapaceum 



* 

* 






* 

* 




P. spathulatum 



* 







* 





P. species (1571) 

* 


* 







* 



* 


P. species (2587) 



* 

* 






* 





P. species (1613) 



* 







* 



* 


P. species (442) 



* 







* 





P. species (2440) 



* 







* 



¥ 


P. species (1456) 



* 







* 



* 


P. species (1653) 

* 


* 







* 





P. species (2439) 



* 







* 

* 




P. species (2443) 



* 







* 



* 


P. species (1454) 



* 












Section 















Sevmouria 















P. asarifolium 









* 

* 

* 




P. dipetalum 



* 







* 





P. marginatum 




* 







* 




P. trifoliatum 



* 

* 












Taxonomic value of trichomes in Pelargonium 


229 


Species 


Section: 
Polyactium 
P. anethifolium 
P. bowkeri 
P. caffrum 
P. flabellifolium 
P. heracleifolium 
P. lobatum 
P. luridum 
P. multiradiatum 
P. pillansii 
P. pulverulentum 
P. radulaefolium 
P. schizopetalum 
P. triste 
P. woodii 

Section: 

Otidia 
P. alternans 
P. carnosum 
P. ceratophyllum 
P. crithmifolium 
P. dasyphyllum 


Petiole 

Non-Glandular Hairs 

Glandular Hairs 

la 

lb 

2a 

2b 

3 

4 

5a 

5b 

6 

7 

8 

9 

10 

11 










* 





* 











* 

* 







* 


* 


* 






* 


* 






* 



* 







* 




* 

* 





* 


* 






* 



* 


* 

* 








* 





* 



* 






* 














* 










* 

* 



* 


* 

* 











* 


* 





* 

* 






* 



* 



* 


* 






* 


* 

* 











* 





* 









* 





* 









* 





* 









* 





* 








1 * 






3J— * 


230 


Journal of South African Botany 


Species 

Lamina 

Non-Glandular Hairs 

Glandular Hairs 


la 

lb 

2a 

2b 

3 

4 

5a 

5b 

6 

7 

8 

9 

10 

11 

Section: 















Polyactium 















P. anethifolium 










* 





P. bowkeri 

* 


* 







* 



* 


P. caffrum 







* 



* 





P. flabellifolium 




* 






* 



* 


P. heracleifolium 



* 







* 

* 

* 



P. lobatum 

* 

* 



* 





* 





P. luridum 




* 






* 



* 


P. multiradiatum 

* 



* 






* 





P. pillansii 




* 






* 





P. pulverulentum 










* 

* 




P. radulaefolium 







* 

* 


* 



* 


P. schizopetalum 



* 







* 





P. triste 



* 

* 






* 





P. woodii 



* 







* 



* 


Section: 















Otidia 















P. alternans 



* 

* 






* 





P. carnosum 



* 

* 






* 



* 


P. ceratophyllum 

* 









* 





P. crithmifolium 



* 







¥ 





P. dasyphyllum 

* 















Taxonomic value of trichomes in Pelargonium 


231 









Petiole 






Species 






























Non-Glandular Hairs 


Glandular Hairs 


la 

lb 

2a 

2b 

3 

4 

5a 

5b 

6 

7 

8 

9 

10 

11 

Section: 















Ligularia 
P. abrotanifolium 






* 





* 


* 


P. aridum 







* 



* 

* 




P. artemisiaefolium 

* 




* 





* 





P. articulation 


* 



* 





* 

* 




P. dolomiticum 







* 

* 


* 



* 


P. exstipulatum 
P. fragile 

* 




* 





* 

* 



* 

P. grandicalcaratum 

* 






* 








P. hirtum 




* 


* 




* 





P. hystrix 




* 






* 



¥ 


P. ionidiflorum 




* 


* 




* 



* 


P. karooicum 

* 














P. oreophilum 

* 



* 






* 



* 


P. ovato-stipulatum 





* 





* 


* 

* 


P. praemorsum 






'* 




* 

* 




P. pulchellum 
P. ramosissimum 

* 




* 



* 


* 




* 

P. sericifolium 



* 












P. xerophyton 















Section: 
Jenkinsonia 
P. antidysentericum 

* 






* 





* 

* 

* 

P. tetragonum 

* 















232 


Journal of South African Botany 


Species 

Lamina 

Non-Glandular Hairs 

Glandular Hairs 


la 

lb 

2a 

2b 

3 

4 

5a 

5b 

6 

7 

8 

9 

10 

11 

Section: 















Ligularia 















P. abrotanifoliwn 






* 





* 


* 


P. aridum 




* 






* 





P. artemisiaefolium 





* 





* 





P. articulatum 




* 






* 





P. dolomiticum 

* 




* 

* 

* 



* 



* 


P. exstipulatum 






* 





* 




P. fragile 

* 




* 





* 




* 

P. grandicalcaratum 



* 







* 

* 




P. hirtum 

* 



* 






* 





P. hystrix 




* 






* 



* 


P. ionidiflorum 

* 



* 






* 



* 


P. karooicum 



* 







* 

* 




P. oreophilum 

* 



* 






* 

* 


* 


P. ovato-stipulatum 

* 









* 

* 


* 


P. praemorsum 

* 









* 

* 




P. pulchellum 



* 



* 




* 



* 


P. ramosissimum 





* 





* 




* 

P. sericifolium 



* 







* 





P. xerophyton 

* 









* 





Section: 















Jenkinsonia 















P. antidysentericum 

* 



* 








* 

. * 

* 

P. tetragonum 
















Taxonomic value of trichomes in Pelargonium 


233 


Species 







Petiole 









Non-Glandular Hairs 



Glandular Hairs 


la 

lb 

2a 

2b 

3 

4 

5a 

5b 

6 

7 

8 

9 

10 

11 

Section: 

Myrrhidium 
P. candicans 
P. longicaule 
-subsp. angusti- 
petalum 
-subsp. caucali- 
folium 

-subsp. convol- 
vulifolium 
-subsp. longicaule 
P. multicaule 
-subsp. trian- 
gular e 

P. myrrhifolium 
-subsp. coriandri- 
folium 

-subsp. myrrhi- 
folium 

P. suburbanum 
-subsp. bipinnati- 
fidum 

-subsp. suburba- 
num 

Section: 

Peristera 

P. chamaedrifolium 
P. fumaroides 
P. grossularoides 
P. harveyanum . 

P. iocastum 

* 

* 

* 

* 

* 

* 

* 

* 

* 

* 

* 

* 

* 

* 


* 

* 

* 

* 

* 

* 

* 

* 

* 

* 

* 

* 

* 

* 

* 


* 


* 


234 


Journal of South African Botany 


Species 


Section: 

Myrrhidium 
P. candicans 
P. longicaule 
-subsp. angusti- 
petalum 
-subsp. caucali- 
folium 

-subsp. convol- 
vulifolium 
-subsp. longicaule 
P. multicaule 
-subsp. trian- 
gulare 

P. myrrhifolium 
-subsp. coriandri- 
folium 

-subsp. myrrhi- 
folium 

P. suburbanum 
-subsp. bipinnati- 
fidum 

-subsp. suburba- 
num 

Section: 

Peristera 

P. chamaedrifolium 
P. fumaroides 
P. grossularoides 
P. harveyanum 
P. iocastum 


Lamina 

Non-Glandular Hairs 

Glandular Hairs 

la 

lb 

2a 

2b 

3 

4 

5a 

5b 

6 

7 

8 

9 

10 

11 



* 







* 





* 



* 






* 





* 









* 





* 


* 







* 





* 









* 





* 

* 





* 



* 





* 

* 



* 





* 





* 

* 



* 





* 




* 

* 


* 


* 


* 


* 

* 





* 



* 





* 

* 





* 









* 





* 









* 





* 









* 





* 



* 






* 



* 


* 


* 







* 






Taxonomic value of trichomes in Pelargonium 


235 


Species 







Petiole 









Non-Glandular Hairs 



Glandular Hairs 


la 

lb 

2a 

2b 

3 

4 

5a 

5b 

6 

7 

8 

9 

10 

11 

Section: 

Campylia 
P. caespitosum 
-subsp. caespito- 
sum 

-subsp. concavum 
P. capillare 
P. coronopifolium 
-subsp. angustissi- 
mum 

-subsp. coronopi- 
folium 
P. elegans 
P. oenotherae 
P. ovale 
P. tricolor 

Section: 

Dibrachya 
P. lateripes 
P. peltatum 

Section: 

Eumorpha 
P. alchemilloides 
P. elongatum 
P. grandiflorum 
P. patulum 
P. quinquilobatum 
P. tabulare 
P. transvaalense 

* 

* 


* 

* 

* 

* 


* 



* 

* 

* 

* 

* 

* 

* 

* 

* 

* 


* 

* 



236 


Journal of South African Botany 


Species 


Section: 

Campylia 
P. caespitosum 
-subsp. caespito- 
sum 

-subsp. concavum 
P. capillare 
P. coronopifolium 
-subsp. angustissi- 
mum 
-subsp. coronopi- 
folium 
P. elegans 
P. oenotherae 
P. ovale 
P. tricolor 

Section: 

Dibrachya 
P. lateripes 
P. peltatum 

Section: 

Eumorpha 
P. alchemilloides 
P. elongatum 
P. grandiflorum 
P. patulum 
P. quinquilobatum 
P. tabulare 
P. transvaalense 


Lamina 

Non-Glandular Hairs 

Glandular Hairs 

la 

lb 

2a 

2b 

3 

4 

5a 

5b 

6 

7 

8 

9 

10 

11 











* 


* 


* 









* 







* 

* 






* 





* 










* 




* 










* 




* 



* 






* 






* 

* 







* 


* 



* 









* 





* 

* 








* 





* 


* 







* 







* 

* 













* 


* 





* 





* 














* 




* 





* 








* 






* 





* 




* 





* 







* 







* 






Taxonomic value of trichomes in Pelargonium 


237 


Species 

Petiole 

Non-Glandular Hairs 

Glandular Hairs 


la 

lb 

2a 

2b 

3 

4 

5a 

5b 

6 

7 

8 

9 

10 

11 

Section: 















Glaucophyllum 















P. laevigatum 





* 










P. lanceolatum 





* 










P. otaviense 










* 

¥ 




P. spinosum 










* 





P. ternatum 

* 














Section: 















Ciconium 















P. acetosum 










* 





P. acraeum 



* 

* 






* 



* 


P. frutetorum 



* 




* 






* 


P. inquinans 



* 

* 






* 





P. zonale 















Section: 















Cortusina 















P. alpinum 






* 

* 








P. cradockense 






* 









P. crassicaule 






* 









P. cortusaefolium 






* 


* 







P. odoratissimum 






* 

* 








P. magenteum 






* 

* 



* 





P. sidaefolium 
















238 


Journal of South African Botany 


Species 

Lamina 

Non-Glandular Hairs 

Glandular Hairs 


la 

lb 

2a 

2b 

3 

4 

5a 

5b 

6 

7 

8 

9 

10 

11 

Section: 















Glaucophyllum 















P. laevigatum 





* 





* 





P. lanceolatum 





* 










P. otaviense 











* 




P. spinosum 










* 

* 




P. ternatum 





* 










Section: 















Ciconium 















P. acetosum 










* 





P. acraeum 







* 

* 


* 





P. frutetorum 



* 

* 






* 



* 


P. inquinans 



* 







* 



* 


P. zonale 

* 


* 







* 





Section: 















Cortusina 















P. alpinum 




* 


* 

* 



* 





P. cradockense 






* 



* 

* 





P. crassicaule 



* 







* 





P. cortusaefolium 



* 





* 

* 




* 


P. odoratissimum 






* 




* 



* 


P. magenteum 







* 


* 

* 





P. sidaefolium 






* 



* 

* 






Taxonomic value of trichomes in Pelargonium 


239 


Species 

Petiole 

Non-Glandular Hairs 

Glandular Hairs 


la 

lb 

2a 

2b 

3 

4 

5a 

5b 

6 

7 

8 

9 

10 

11 

Section: 















Pelargonium 















P. englerianum 




* 






* 





P. betulinum 










* 



* 


P. capitatum 







* 

* 


* 





P. cordifolium 

* 






* 





* 

* 


P. crispum 

* 


* 

* 









* 


P. cucullatum 

* 






* 


* 

* 


* 

* 


P. denticulatum 





* 

* 




* 





P. glutinosum 





* 






* 




P. hermanniae- 















folium 

* 









* 





P. hispidum 

* 














P. papilionaceum 




* 











P. quercifolium 

* 














P. radens 

* 

* 













P. ribifolium 

* 

* 













P. scabroide 




* 











P. scabrum 

* 



* 

* 





* 





P. semitrilobum 

* 



* 











P. sublignosum 

* 












* 


P. tomentosum 















P. vitifolium 






* 



L 







240 


Journal of South African Botany 


Species 

Lamina 

Non-Glandular Hairs 

Glandular Hairs 


la 

lb 

2a 

2b 

3 

4 

5a 

5b 

6 

7 

8 

9 

10 

11 

Section: 















Pelargonium 















P. englerianum 

* 




* 





* 

* 


* 


P. betulinum 

* 

* 



* 





* 





P. capitatum 



* 

* 



* 

* 


* 


* 

* 


P. cordifolium 

* 









* 





P. crisp am 

* 




* 





* 

* 


* 


P. cucullatum 

* 






* 



* 


* 

* 


P. denticulatum 





* 






* 


* 


P. glutinosum 





* 






* 




P. hermanniae- 















folium 

* 




* 





* 





P. hispidum 

* 


* 

* 






* 

* 


* 


P. papilionaceum 




* 






* 


* 

* 


P. quercifolium 

* 



* 


* 

* 

* 


* 



* 


P. radens 

* 




* 





* 

* 


* 


P. ribifolium 

* 

* 








* 



* 


P. scabroide 





* 





* 





P. scabrum 


* 



* 





* 

* 


* 


P. semitrilobum 

* 









* 





P. sublignosum 


* 



* 





* 

* 


* 


P. tomentosum 

* 



* 



* 


* 

* 



* 


P. vitifolium 



* 

* 



* 

* 


* 


* 

* 



Discussion 

In the genus Pelargonium the indumentum of the leaves is generally of a 
hirsute to hispid type, with a pilose covering relatively rare and woolliness 
occurring only in some taxa. Glandular hairs are mostly small, with long 
glandular hairs occurring more prominently in certain sections. In some 
cases a specific trichome type is present only in a few species of a section. In 
such cases it can be used to distinguish between species, e.g. P. suburbanum 
is the only species with curly hairs in the section Myrrhidium. 

If a specific indumentum type occurs in most of the species of a section, 
that indumentum type can be regarded as characteristic of the section; e.g., 
the petioles and laminae of species in the section Hoarea are mostly strigose, 
those in the sections Pelargonium, Glaucophyllum and Eumorpha hispid to 
setose, in the section Cortusina pilose and in the section Otidia the indumen- 
tum is generally sericeous. 

Long glandular hairs occur especially in the sections Hoarea, Ciconium, 
Pelargonium and Polyactium , and are lacking in the sections Seymouria, 


Taxonomic value of trichomes in Pelargonium 


241 


Otidia (except P. carnosum (L.) L’Herit.), Myrrhidium, Peristera (except 
P. harveyanum Schltr. ex Knuth), Dibrachya and Glaucophyllum. Glandu- 
lar hairs with pear-shaped heads occur in the sections Ligularia, Jenkinsonia 
and Myrrhidium. Glandular hairs are completely absent on the leaves of P. 
laevigatum (L.f.) Willd. and P. lancolatum (Cav.) Kern, (section Glauco- 
phyllum ), P. grandiflorum (Andr.) Willd. (section Eumorpha) and P. lat- 
eripes L'Herit. (section Dibrachya). In some species of the sections Polyacti- 
um , Peristera and Glaucophyllum non-glandular hairs are completely absent 
and only glandular hairs occur in these cases. In the section Glaucophyllum 
the indumentum type is very conspicuous with only small glandular hairs 
and characteristic stiff, spiny to bristly non-glandular hairs present. In some 
species of especially the sections Polyactium , Dibrachya , Eumorpha and 
Glaucophyllum , as well as P. tetragonum (L.f.) L'Herit. (section Jenkinso- 
nia) and P. spathulatum (And.) DC. (section Hoarea), the lamina and/or pe- 
tiole is glabrous or nearly so. In some sections, i.e. Polyactium, Eigularia, 
Dibrachya , Eumorpha, Ciconium and Pelargonium the indumentum occurs 
only, or more densely, in the adaxial groove of the petiole. 


Conclusions 

Due to the fact that in every species the trichome complement consists of 
a number of characteristic trichome types, and that specific trichome types 
occur only in certain sections, while in the different species and subspecies 
of each section the trichomes themselves also differ, trichomes can be of tax- 
onomic value on section, specific and infra-specific level in the genus Pelar- 
gonium. 

Despite the interspecific variation in trichome density, the indumentum 
type is often characteristic of specific sections (vide Discussion). In the sec- 
tions the trichome complement consists of certain combinations of trichome 
types which differ from the combinations found in other sections (Table 1). 
Within the sections species can also often be grouped together according to 
their trichome complement. This is probably an indication of a close re- 
lationship among these species. In the section Myrrhidium, for example, the 
difference between P. longicaule Jacq., P. myrrhifolium (L.) L Herit. and 
P. suburbanum Clifford, and the interrelationship of their subspecies are 
clearly reflected in the similar trichome complement of the subspecies of 
every species. The restriction of glandular hairs with pear-shaped heads to 
the sections Ligularia, Jenkinsonia and Myrrhidium confirms the presumed 
relationship of these sections. 

The occurrence of a specific trichome type on one organ does not neces 
sarily imply that it occurs on the other organs. The glandular haiis with 
pear-shaped heads were found on both the petiole and lamina of on v 


242 


Journal of South African Botany 


P. ramosissimum (Cav.) Willd. and P. fragile (Andr.) Willd. (section Ligu- 
laria), P. antidysentericum (Eckl. & Zeyh.) Kostel. (section Jenkinsonia) 
and P. myrrhifolium (L.) L’Herit. subspecies myrrhifolium (section Myrrhi- 
dium). According to Marais (1980) however, this trichome type occurs on 
the floral parts of all species in the sections Jenkinsonia and Myrrhidium, 
and in the case of P. tetragonum (L.f.) L’Herit. (section Jenkinsonia ), where 
it does not occur on the leaf, it is mentioned by Fourie (1978) that this type 
occurs on the stem. 

References 

Carlquist, S., 1961. Comparative plant anatomy. New York: Holt, Rinehart & 
Winston. 

Fourie, L. J., 1978. Vergelykende embriologie en anatomie van Pelargonium pelta- 
tum (L.) L’Herit. en P. tetragonum (L.f.) L'Herit. M.Sc. thesis. University 
of Port Elizabeth. 

Johnson, H. B., 1975. Plant pubescence: an ecological perspective. Bot. Rev. 41,3: 
233-258. 

Marais, E. M., 1980. Die taksonomiese waarde van die blom en stuifmeel van Pelar- 
gonium. M.Sc. thesis. University of Stellenbosch. 

Phillips, E. P., 1951. The genera of South Africa flowering plants. Pretoria: Govern- 
ment Printer. 

Van der Walt, J. J. A., 1977. Pelargoniums of Southern Africa. Cape Town: Pur- 
nell & Sons. 

Willis, J. C., 1973. A dictionary of the flowering plants and ferns. Cambridge: Uni- 
versity Press. 


Ji S. Afr. Bot. 49 (3): 243-249 (1983) 


A NEW SPECIES OF BOOPHANE HERBERT (AMARYLLIDACEAE) 
FROM THE NORTH WESTERN CAPE 


Deirdre Snijman 

(Compton Herbarium, Kirstenbosch, Private Bag X7, Claremont 7735 
R.S.A.) 

Abstract 

A new yellow-flowered species of Boophane from the north western Cape, is de- 
scribed and illustrated. 

UlTTREKSEL 

'N NUWE SOORT BOOPHANE HERBERT (AMARYLLIDACEAE) VANAF 
DIE NOORDWESTELIKE-KAAP 

’n Nuwe Boophane - spesie met geel blomme uit die noordwestelike Kaapprovinsie 
word beskryf en geillustreer. 


Key words: Boophane, sp. nov., Amaryllidaceae, north western Cape. 


Boophane is a small genus of the Amaryllidaceae distributed throughout 
eastern and southern Africa. In the Flora Capensis (1896) Baker recognised 
three species. Two additional species, B. haemanthoides Leighton and 
B. pulchra Barker, both from the western Cape, were described in 1947 and 
1963 respectively. At the time of their publication Leighton and Barker col- 
lected several more specimens of Boophane from Namaqualand, unlike any 
thus far known. To these specimens (at Kew and the Compton Herbarium) 
Miss Barker appended the unpublished name B. fiava. 

Field work in Namaqualand during the past three years yielded confirma- 
tion of the distinctiveness of B. fiava. The species is described here for the 
first time. 

Boophane fiava Barker ex Snijman, sp. nov. 

Boophane fiava Barker ex Snijman; species floribus flavis et staminibus 
declinatis congeneribus diversa. 

Bulbus profunde hypogaeus, anguste ad late ovatus, 100-130 mm diam., 
tunicis crustaceis ferrugineis, superne in collum productis; collum transverse 


Accepted for publication 21st February, 1983. 

243 


244 


Journal of South African Botany 


porcatus. Folia (3-)4-6, hysterantha, prostrata, laminis late loratis, 210— 
420 mm longa, 50-110 mm lata, plana, subtus rubropunctata, margine cilia- 
to. Pedunculus complanatus, 200-280 mm longus, 20-50 mm latus, pallidus 
flavovirens. Spathae valvae 2, oblongae, mox marcescens. Umbella globosa, 
200-400 mm lata. Flores 40-220, parvi, flavi, ferrugineis apice. Pedicili 
rigidi, 70-200 mm longi, trigoni. Perianthium tubo 1-2 mm longo, segmentis 
lanceolatis reflexis 9-14 mm longis et 2 mm latis, undulatis prope basin. 
Filamenta declinata, 8-12 mm longa, tumore basi, marronina apicem versus. 
Anthera 2 mm longa, marronina. Ovarium trigonum, 3 mm latum, viridis, 
demum rubescens. Stylus plus minusve rectus, 5-8 mm longus. Capsula tri- 
gona, membranacea, fere 10 mm lata. Semina globosa carnosa, fere 5 mm 
diam. 

Type: South Africa, Cape Province, 20 km south of Karkams, fl. 17/4/1981, leaves 
ex hort. Kirstenbosch 9/9/1981, Snijman 438 (NBG, holotype; K, PRE, isotypes). 

Figs 2 & 3. 

Bulb deep seated, narrowly to broadly ovate, 100-130 mm diam.; outer 
tunics hard and brittle, rust-coloured, splitting ultimately into numerous ver- 
tical segments with raised transverse bands forming a neck. Leaves (3-)4-6, 
appearing after the inflorescence, recurved and spreading, finally prostrate; 
blade broadly lorate, 210-420 mm long, 50-110 mm broad, flat, light green, 
leathery; abaxial surface marked with fine angular red speckles; margin with 
a white to rusty, soft or chaffy fringe. Peduncle flattened, 200-280 mm 
long, 20-50 mm across greatest diam., pale yellowish-green, glabrous. 
Spathe valves 2, oblong, 40-75 mm long, 20-50 mm wide, yellowish-green, 
soon becoming withered and reflexed. Umbel large, globose, 200-400 mm 
across. Flowers 40-220, small, pale yellow, becoming light brown at the tips. 
Pedicels rigid, 70-200 mm long, ageing to pale maroon, trigonous. Perianth 
tube 1-2 mm long; segments reflexed, lanceolate, 9-14 mm long, 2 mm 
wide, usually undulate near the base; tips acute. Filaments declinate, 8- 
12 mm, with a large swelling at the base, yellow basally, maroon distally. 
Anthers 2 mm long when open, maroon with yellow pollen. Ovary tri- 
gonous, 3 mm across, dark green to reddish. Style more or less straight, 
about 5-8 mm long, yellowish. Capsule trigonous, membranous, about 10 
mm across. Seeds globose, fleshy, about 5 mm diam. 

Flowering time: March to May. 

Leafing period: from May until September. 

Distribution: from Komaggas in the north, through Namaqualand to the 
Bokkeveldberge and the Gifberge in the south, at elevations ranging from 
475 to 750 metres above sea level. The species favours flat terrain but shows 
no substrate preference. In the north the soils are coarse and granitic and 


New species of Boophane Herbert (Amaryllidaceae) 


245 



Fig. 1. 

A: Boophane flava Barker ex Snijman, growing in Namaqualand Broken Veld neai 
Karkams; B: detail of inflorescence; C: detail of bulb showing transversely thickened 

bands on neck. 


BJ — 5 


246 


Journal of South African Botany 





A Aim, 


anno 1981-04-17 
Leaves ex hort 
ACr 1981-09-09 


Boophane flava Barker toe. 

south of Karkams. East slope in 
coarse granite soil, open terrain. Six 
plants in population, all flowering. 
Peduncle and pedicels greenish-yellow* 
Flowers yellow with brown tips .Filaments 
maroon, yellow at the base. Anthers 
maroon. Leaves six, flat on ground* 
Kargin of soft brown hairs. Undersurface 
of leaves speckled with red. 


Fig. 2. 

Type collection of Boophane flava Barker ex Snijman with inflorescence in Compton 

Herbarium. 



New species of Boophane Herbert (Amaryllidaceae) 


247 




NATIONAL BOTANIC (.ARDENS OF 
SOUTH AFRICA 
KIRSTENBOSCH. StVflANDS. C'.P. 


3017 DB GRiO WEf 

(Hondeklipbaai) 

D. Snijman 

U38 no 

RSGlO CAPE 

1S81-0U-17 
Leaves ex hort 
Al - T 1981-09-09 

Boophane flava Barker ms. 


20km south of Karkams. East slope in 
coarse granite soil, open terrain. Six 
plants in population, all flowering. 
Peduncle and pedicels greenish-yellow. 
Flowers yellow with brown tips. Filaments 
roon, yellow at the base. Anthers 
aaroon. Leaves six, flat on ground. 14a rg in 
of soft brown hairs. Undersurface of 
leaves speckled wl th red. 


hto-JLi- A^vJL ^ LkaJo. i>a*kjs jUc 

Jb Aa*j****J 

Fig. 3. 


no. j. 

Type collection of Boophane flava Barker ex Snijman with leaves, peduncles, neck 
fibres and floral parts in Compton Herbarium. 



248 


Journal of South African Botany 


the cover is mostly of low succulent bushes. On the Bokkeveldberge the 
soils are heavy clays, whereas in the south the species grows in deep sand of 
Table Mountain Sandstone origin in association with arid fynbos. Popu- 
lations are localised made up of more or less ten scattered individuals (Fig. 

4). 

Boophane flava is most distinctive. Unlike any other species of Boophane 



it has small yellow flowers with declinate stamens and a bulb with a well- 
developed neck made up of frayed vertical segments with transversely thick- 
ened bands. Boophane guttata (L.) Herbert, previously known as B. ciliaris 
(L.) Herbert from the south western Cape, is its closest ally. The latter also 
has prostrate ciliate leaves and small flowers but differs in that the flowers 
are actinomorphic and maroon-coloured. 

Bulbs of B. flava in cultivation at Kirstenbosch for the past eight years 
have never flowered. The habit of flowering irregularly and sometimes only 
after long intervals, seems to apply to bulbs in nature as well. The good 
autumn rain of 1981 caused B. flava to flower profusely in the Grootvlei and 
Kamieskroon areas, sometimes even in private gardens where previously the 
flowers had never been seen. 

The flowers are mostly without a scent but produce copious nectar. Dur- 
ing the day ants, bees and butterflies were seen visiting the flowers of which 
the bees appeared to be the only effective pollinators. The pale yellowish- 
green colour of the inflorescences made them appear most conspicuous at 
dusk when the flowers may also be visited by moths. 


New species of Boophane Herbert (Amaryllidaceae) 


249 


CAPE — 2917 (Springbok): near Van Reenen se Water, Kamaggas 200 (-DA), van 
Berkel 371 (NBG); Komaggas (-DC), Herre 1811 (BOL). 

— 3017 (Hondeklipbaai): Schulpfontein (-AA), van Berkel 403 (NBG); Grootvlei 
(-BB), Leighton sub NBG 265/45 (NBG), Snijman 440 (NBG, PRE), van Berkel 319 
(NBG); farm Brakdam (-BD), Snijman 439 (K, MO, NBG, PRE); just north of 
Darter's Grave, Snijman 604 (NBG, PRE); Kharkams, Barker 9869 (K, NBG); 
20 km south of Karkams (-DB), Snijman 438 (K, NBG, PRE). 

— 3118 (Vanrhynsdorp): farm Koelfontein, south of Rietpoort (-AA), Snijman 705 
(NBG); top of Koebee Pass (-DB), Perry 1552 (NBG); Gifberg Plateau (-DC), W. 
Olivier 192 (NBG); Giftberg, Herre sub SUG 67 (BOL). 

— 3119 (Calvinia): farm Glen Lyon, east of Nieuwoudtville (-AC), Snijman 602 
(NBG). 

Acknowledgments 

I am sincerely grateful to Nicky and Fred van Berkel for so freely sharing 
their knowledge of Namaqualand; to my parents for their invaluable assist- 
ance with the gathering of the deeply lodged bulbs and to Graham Duncan 
for his ever-willing help with cultivating the bulbs. 

References 

Baker. J. G.. 1896. In: Thiselton-Dyer, W. T. (ed.). Flora Capensis 6: 242-244. 
Ashford, Kent: Reeve & Co. 

Leighton, F. M., 1947. Plantae Novae Africanae. Jl S. Afr. Bot. 13: 59-61. 

Barker, W. F., 1963. Two new species of Amaryllidaceae. Jl S. Afr. Bot. 29: 
163-165. 






J1 S. Afr. Bot. 49 (3): 251-259 (1983) 


NOTES ON TULBAGHIA : 5. SCANNING ELECTRON MICROSCOPY OF 
SEED-COAT PATTERNS IN NINETEEN SPECIES* 

Canio G. Vosa 

( Botany School, South Parks Road, Oxford 0X1 3RA, England) 

Abstract 

Scanning Electron Microscopy analysis of seed-coat patterns of nineteen species 
of Tulbaghia has revealed specific differences between them. The differences include 
variation in the size, form and sculpturing of the seed-coat surface cells. Each species 
possesses its own constant micro-morphological characteristics which can be used as 
an additional taxonomic marker. 

UlTTREKSEL 

AANTEKENIN GE OOR TULBAGHIA : 5. SKANDEER-ELEKTRONMIKRO- 
SKOOP ONDERSOEK VAN DIE PATRONE OP DIE SAADHUID VAN 
NEGENTIEN SOORTE 

Skandeer-elektronmikroskoop ondersoek van saadhuid patrone van negentien 
Tulbaghia - soorte het spesifieke verskille aan die lig gebring. Die verskille sluit vari- 
asie in grootte, vorm van die oppervlakkige saadhuid-selle in. Elke soort het 
konstante mikro-morfologiese kenmerke wat as bykomende taksonomiese merker 
gebruik kan word. 

Key Words: Tulbaghia, Scanning Electron Microscopy, testa micro-morphology. 
Introduction 

Most Tulbaghia species are notoriously difficult to identify as herbarium 
specimens (Vosa, 1975; Burbidge, 1978). Certain fine details of morphology, 
such as the fleshiness of the corona, are usually poorly preserved in dried 
material. However, ripe seeds are sometimes found on herbarium specimens 
and with this in mind, the present study concerns the use of Scanning Elec- 
tron Microscopy of the seed-coat of Tulbaghia species. It includes all the 
species recognized by Vosa (1975) with the exception of T. cameroni Bak. 
and T. rhodesica Fries, for which no adequate seed samples were available. 

Material and Methods 

Ripe seed samples were collected from correctly identified species both 
on living plants and on herbarium specimens. Seeds from three to five 

* Continued from J1 S. Afr. Bot. 45: 127-132 (1979); 46: 109-114 (1980): 
47: 57-61 (1981): 48: 241-244 (1982). 

Accepted for publication 28th March, 1983. 

251 


252 


Journal of South African Botany 


plants, from different collections of each species, were sputter-coated with 
gold to about 20 nm thickness with a Polaron E-5000 coating unit, and 
examined in a Cambridge Stereoscan S-150 Scanning Electron Microscope. 
No fundamental differences were found in seed-coat patterning in different 
parts of the seed (Fig. 1). 



Fig. 1. 

T. transvaalensis : the entire seed. Note uniformity of seed-coat pattern, x 140 ca. 


Results 

The gross morphology of the seed of Tulbaghia is very uniform. All 
species have black on dark-brown coated seeds, which are typically elong- 
ated and usually very much creased and compressed into a wedge shape with 
a somewhat triangular section (Fig. 1). 

Scanning Electron Microscopy reveals characteristic cell shapes and 
sculpturing of the outer cell wall of the testa. 

No major differences in testa micro-morphology were observed between 
seeds from different collections of the same species. 

Figures 2, 3, 4 and 5 show the differences between the species. 


Notes on Tulbaghia: 5 


253 



Fig. 2. 

a. T. capensis ; b. T. alliacea ; c. T. cernua; d. T. ludwigiana ; e. 7. dregeana ; 
f. T. tenuior, X 500 ca. 



254 


Journal of South African Botany 


Fig. 3. 

a. T. nutans: b. T. macrocarpa : c. T. transvaalensis; d. T. acutiloba: e. T. simmleri : 

f. T. galpini, x 500 ca. 


Notes on Tulbaghia: 5 


255 



T. leucantha (2n = 2x = 12); b. T. leucantha (2n - 4x - 24); c - T. natalensis, 
d. T. verdoornia ; e. T. coddii', f. T. montana , x 5U 


256 


Journal of South African Botany 



— — — i ii ii • 

Fig. 5. 

a. T. violacea from the eastern part of the range; b. T. violacea from the western part 
of the range; c. T. violacea var. maritima ; d. T. cominsii, x 500 ca. 


Descriptions 
Figure 2. 

a. T. capensis L. (C. G. Vosa, 294/14, OXF): clear, distinctly raised 
cells, depressed and deeply pitted in the centre. 

b. T. alliacea L.f. (C. G. Vosa, 61, OXF): flat appearance, cells with rib- 
bon-like sutures and covered with an irregular netting of excrescences. 

c. T. cernua Ave-Lall. (C. G. Vosa, 413/32, OXF and cultivated speci- 
mens): a clear papillate surface with somewhat sharp crowded ridges or 
crests. 


Notes on Tulbaghia: 5 


257 


d. T. ludwigiana Harv. (C. G. Vosa, 398/29, OXF): flattish cells with 
small excrescences and short ridges. 

e. T. dregeana Kunth (C. G. Vosa, 62, OXF): somewhat polyhedric, al- 
most smooth cells, with slightly creased upper surfaces. 

f. T. tenuior Krause & Dinter (C. G.Vosa, 136 OXF): small polygonal 
cells with a crowded network of shallow ridges. 

Figure 3. 

a. T. nutans Vosa (C. G. Vosa, 536/55, OXF): straight-sided cells with a 
flat crinkled surface. 

b. T. macrocarpa Vosa (C. G. Vosa, 132 OXF): round, well separated 
cells with a flat top covered with shallow ridges. 

c. T. transvaalensis Vosa (C. G. Vosa, 486/49, OXF): plump, mostly flat- 
topped cells with minute excrescences and some longish furrows at the 
edges. 

d. T. acutiloba Harv. (C. G. Vosa, 347/21, OXF and cultivated speci- 
mens): irregular, polygonal, flat-topped cells, with shallow pits ringed with 
furrow-like sculpturing at the edges. 

e. T. simmleri Beauv. (C. G. Vosa, 487/50, OXF, and cultivated speci- 
mens): irregular shaped cells with well defined sutures and shallow mostly 
continuous ridges in waving lines. 

f. T. galpini Schl. (C. G. Vosa, 60, OXF): largish, flat cells with well 
defined sutures and covered with irregularly shaped pits. 

Figure 4. 

a. T. leucantha Bak. (C. G. Vosa, 430/35, OXF, 2n = 2x = 12): smooth 
looking cells with a minute vermiculate appearance. 

b. T. leucantha Bak. (C. G. Vosa, 449/38, OXF, 2n = 4x =24): similar 
but somewhat larger and rounder cells than the diploid form. 

c. T. natalensis Bak. (C. G. Vosa, 421/34): very similar to T. leucantha 
but flatter cells with slightly larger vermiculations. 

d. T. verdoornia Vosa & Burbidge = T. carnosa Burbidge (C. G. Vosa, 
1599, OXF): flattish, irregular cells with a minutely pitted surface. 

e. T. coddii Vosa & Burbidge = T. poetica Burbidge (C. G. Vosa 
496/52): flat appearance with mostly elongated cells covered with a network 
of more or less regularly spaced pits and raised, ribbon-like sutures. 

f. T. montana Vosa (Bayliss, 7831, OXF): somewhat polygonal cells with 
definite boundaries and sharpish, regularly spaced excrescences. 

Figure 5. 

a. T. violacea Harv. (C. G. Vosa, 308/17, OXF): flat-topped cells with a 


258 


Journal of South African Botany 


crinkled and ridged surface and well defined depressed sutures. This collec- 
tion is from the eastern part of the range. 

b. T. violacea Harv. (C. G. Vosa, 371/23 OXF): as a. but usually larger 
more irregular cells. This collection is from the western part of the range of 
T. violacea , about 350 km SW of the a. collection. 

c. T. violacea Harv. var. maritima Vosa (C. G. Vosa, 290/12): large, 
elongated cells, welded together by ribbon-like smooth sutures, strikingly 
different from T. violacea sensu strictu. 

d. T. cominsii Vosa (C. G. Vosa, 1568, OXF): irregularly shaped, elong- 
ated cells, covered with a network of small excrescences. Cell sutures rib- 
bon-like, well defined and smooth. 

Except where indicated, the above descriptions refer to diploid plants. 
The hexaploid forms of T. alliacea and T. capensis , as well as the tetraploid 
form of T. cernua (Vosa, 1975), have been investigated and, as in the case 
of the diploid and tetraploid forms of T. leucantha , illustrated in Figure 4, a. 
and b., no fundamental differences, apart from the average cell size, were 
found. 


Discussion 

The differences between T. alliacea (Fig. 1, b.) and T. cernua (Fig. 1, 
c.), already discussed by Vosa (1981, pp. 58-61), are also evident in their 
testa micro-morphology. The flat, well sutured cells of T. alliacea are in fact 
very different from the domed, somewhat bell-shaped cells of T. cernua. 

The recognized morphological and cytological relationship between T. 
leucantha (Fig. 3, a. and b.), T. verdoornia (Fig. 3, d.) and T. natalensis 
(Fig. 3, c.), is confirmed by the similarity of the seed-coat patterns. The fur- 
ther two species in their group (Group 5, Vosa, 1975, pp. 57-60), T. galpini 
(Fig. 2, f.) and T. coddii (Fig. 3, e.) possess a testa micro-morphology which 
is distinct from one another as well as from the other species. 

T. alliacea, T. galpini, T. coddii, T. violacea var. maritima and T. comin- 
sii, though well distinct from each other, share a characteristic testa with a 
flat-cell appearance and well defined ribbon-like sutures, quite unlike other 
Tulbaghia species. 

The existence of such similarity may have an ecological significance in re- 
lation to seed dispersal and germination. In this context it is interesting to 
note that the above five species, with the exception of T. galpini, at least in 
the southern area of its range, normally inhabit humid places which are 
usually periodically flooded. The possession of close cell sutures may play a 
part in delaying water uptake, probably permitting longer distance seed dis- 
persal, which, in Tulbaghia, is effected by the swaying fruiting stems when 
the loculicidal capsule opens. 


Notes on Tulbaghia: 5 


259 


The S.E.M. of the testa micro-morphology of nineteen species of Tul- 
baghia has shown certain unique and species specific features. Apart from 
the ecological significance, these unique features, as well as the similarities, 
can be regarded as a useful addition to the more usual characters used in 
taxonomic identification in Tulbaghia. 

Acknowledgements 

I wish to thank Lindsay Helsby, Gavin Wakley and David Kerr for their 
technical assistance and advice. 

References 

Burbidge, R. B., 1978. A revision of the genus Tulbaghia. Notes R. bot. Gdn Edinb. 
36 : 77-103. 

Vosa, C. G., 1975. The cytotaxonomy of the genus Tulbaghia. Anrtali Bot. (Roma) 
34 : 47-121. 




J1 S. Afr. Bot. 49 (3): 261-304 (1983) 


STUDIES IN CYPERACEAE IN SOUTHERN AFRICA: 10. THE GENUS 
KYLLINGA ROTTB. 

Fiona M. Getliffe 

(Bern Botanical Laboratories, University of Natal, P.O. Box 375, Pieter- 
maritzburg 3200, R.S.A.) 

Abstract 

A review of the taxonomic history of the genus Kyllinga Rottb. is given. Diag- 
nostic criteria are discussed and emphasis laid on the morpology of the underground 
parts as key characters. Fourteen species are described including two new records for 
the area, K. polyphylla Willd. ex Kunth and K. nemoralis (Forst.) Dandy. A key to 
the genus in South Africa is provided. 

UlTTREKSEF 

STUDIES VAN CYPERACEAE IN SUIDELIKE AFRIKA: 10. DIE GENUS 
KYLLINGA ROTTB. 

'n Oorsig van die taksonomiese geskiedenis van die genus Kyllinga Rottb. is ge- 
lewer. Diagnostiese kenmerke word bespreek en nadruk word op die ondergrondse 
dele gele. Viertien spesies word beskryf insluitende twee nuwe rekords vir die ge- 
bied, naamlik K. polyphylla Willd. ex Kunth en K. nemoralis (Forst.) Dandy, n 
Sleutel tot die genus in Suid-Afrika word verskaf. 

Key words: Kyllinga, Cyperaceae, southern Africa 

Generic limits within the Cyperaceae have frequently been subject to dif- 
fering taxonomic interpretations. Kyllinga Rottb. is no exception having 
been treated as a genus (Vahl, 1806; Nees, 1835/6), united with Mariscus 
Gaertn. (Boeckeler, 1868) and placed in the genus Cyperus L. together with 
Mariscus, Pycreus Beauv. and Juncellus C. B. Clarke (Kiikenthal, 1936; 
Koyama, 1961). This latter treatment has not enjoyed universal acceptance 
however, and has been challenged on chemical grounds (Allan et al., 1978). 
In the absence of the unequivocal evidence to support Kiikenthal s decision, 
Kyllinga will be maintained at the generic level in this account. 

The name Kyllinga Rottb. has been conserved in preference to Thryoce- 
phalum J. R. and G. Forst. (1776) despite an earlier homonym Kyllinga 
Adans. (Apiaceae) (Lanjouw, 1978). Rottbbll (1773) described Kyllinga and 
five species of which K. monocephala was the first mentioned. This species 


Accepted for publication 28 March, 1983. 


BJ— 6 


261 


262 


Journal of South African Botany 


has been designated the type species (Lanjouw, l.c.) with the correct name 
K. nemoralis (Forst.) Dandy (Dandy, 1935). 

Vahl (1806) included Central African species in his list of seven Kyllinga 
species, but the earliest record of South African species was given by Nees 
(1835/6) with a description of four new species of Kyllingia, a spelling now 
rejected in favour of Kyllinga. Kunth (1837) recorded twenty-eight species, 
commenting that the genus differed from Pycreus only in the number of 
florets, while Boeckeler (1868) gave a rather broader view by including in 
Kyllinga the genus Mariscus with its three style branches where Kyllinga , in 
the strictest sense, has two. Bentham and Hooker (1883) and Engler and 
Prantl (1889) restored Mariscus to generic status. 

The first comprehensive account of the genus in southern Africa, how- 
ever, was that of Clarke (1897) which followed his publication of a list of 
twenty-one African species (1895) and his contribution of six species of Afri- 
can and Indian origin (1893). In his synopsis of the genus, Clarke (1897) re- 
cognised three subgenera; Pseudo-Pycreus, Eukyllinga and Thryocephalum, 
number of nutlets and presence or absence of a wing on the glumes being 
the diagnostic characteristics. Ten species were recognised but in 1901, 
Clarke presented a second major account in which thirty-eight tropical 
species were recognised. This was followed by a species list (1908) and illus- 
trations (1909). 

Kyllinga was included in accounts by Schonland (1922) and Brain (1931); 
but the next major study of the genus was that of Kiikenthal (1936) in which 
the genus Kyllinga was reduced to a subgenus of Cyperus together with 
Mariscus, Torulinium, Juncellus, Pycreus and Eucyperus. The subgenus Kyl- 
linga sensu Kiikenthal, was further subdivided into sections reflecting 
Clarke’s classification (1897) and within section Eukyllinga , three groups 
were defined on the basis of rhizome characteristics. The arrangement of the 
three sections suggests an evolutionary progression involving a reduction in 
number of florets per spikelets and the development of a wing on the keel of 
the fertile glume. 

Koyama (1961) followed Kiikenthal adding only the genus Lipocarpha 
R. Br. to the list of those sunk into Cyperus sensu Kiikenthal. Kern (1974) 
also included Kyllinga within Cyperus, but other authors (Podlech, 1960 and 
Hooper, 1972) do not agree, preferring to retain Kyllinga as a separate 
genus. This latter policy is adopted here. 

The genus Kyllinga Rottb. is an essentially African genus, but three 
species are recorded from tropical America and India, four endemic species 
from the West Indies and seven endemic to the Malagasy Republic. Four- 
teen species are represented in southern Africa. 

Members of the genus may be recognised in the field by their single, 
compact inflorescence and distinguished from similar forms in Mariscus by 


Studies in Cyperaceae in southern Africa: 10 


263 


their paired style branches. The underground organs, in particular, have a 
characteristic smell faintly reminiscent of lemon. Each spikelet comprises a 
pair of sterile glumes (the lower bract and prophyll) subtending distichously 
arranged fertile glumes. The lowest floret is bisexual, consisting of an 
ovary with two style branches and three (or 2 or 1) stamens. The upper 
floret may be male or, rarely, bisexual. Most species produce a single nutlet 
from the fertile bisexual flower but K. pulchella Kunth has 2-4 bisexual 
flowers and may produce 2 nutlets. There are no hypogynous bristles. 

Diagnostic Criteria 

Emphasis has been placed on spikelet form, number of nuts, the wing on 
the keel of the glume and the perennating organs. Each of these was studied 
and will be discussed together with other characters found to be of interest. 

Kyllinga is distinguished from neighbouring genera by its single compact 
inflorescence of sessile spikelets. In general, this is a consistent feature of 
the genus though species such as K. polyphylla Willd. ex Kunth and K. ela- 
tior Kunth frequently have minor sessile heads in the axils of lower inflores- 
cence bracts. Two aberrant collections have been made; one a triple head in 
K. alba Nees and one collection of K. alata Nees with two peduncled, minor 
heads reminiscent of Mariscus. These are, however, not consistent features 
for these species. 



Fig. 1 

Anomalous inflorescences of Kyllinga species. A. K. alata from the Bluff grasslands, 
Durban. B. K. alba from the Lebombo Mountains. 


The shape of the inflorescence is useful in separating species. K. elatior 
and K. odorata Vahl invariably have cylindrical heads while all the other 
species have spheroidal heads. Colour is also useful if used with caution. K. 
alata invariably has golden spikelets with green keels giving a characteristic 
golden inflorescence and K. pulchella has deep red inflorescences ■ o orata 



264 


Journal of South African Botany 


inflorescences are white when mature, as are those of K. alba. In K. alba 
however, some discolouration is possible with cinnamon spots appearing. 
Associated with the inflorescence are leafy bracts which are useful key 
characters in some species. The orientation of three stiff bracts, erect at first 
and later reflexing to a 45° angle, is characteristic of K. erecta Schum., K. 
intricata Cherm., K. brevifolia Rottb., K. melanosperma Nees, and K. pauci- 
flora Ridl.. K. elatior is easily recognised in the field by its long, horizon- 
tally-orientated bracts, but this is not always as apparent in herbarium ma- 
terial and rhizome and inflorescence shape are more reliable here. 

Spikelet features have been used as diagnostic features at the subgeneric 
level and to distinguish species. One of these features was the number of bi- 
sexual flowers and thus the number of nutlets per spikelet. In these respect, 
K. pulchella is unique amongst the southern African species in bearing 2-4 
(as opposed to one) fertile floret per spikelet. 

Kyllinga florets usually include three stamens, but three South African 
species differ from the norm. K. brevifolia invariably has one stamen per 
floret and K. odorata two. K. welwitschii Ridley has one or two. 

The presence of a wing on the keel of the fertile glumes has been used to 
define the subgenus Thryocephalum (Clarke, 1901) and particular attention 
has been given to this characteristic. Within a population variation may be 
very slight, but comparisons of individuals from different populations over 
the range of distribution of K. alba and K. alata showed considerable varia- 
tion (see Figure 2). As a diagnostic criterion some caution must be exercised 
when using this characteristic. In the present study, wing width was found to 
be a function of the age of the spikelets of K. alba and thus is not a useful 
diagnostic or key character when dealing with herbarium specimens of this 
species. 

As with many members of the Cyperaceae, the rhizome or other peren- 
nating underground stems have proved to be invaluable aids to identifica- 
tion. Indeed, with experience of the genus it is usually possible to make ac- 
curate specific identifications of vegetative material on this basis alone. 

The term “rhizome” has been used somewhat loosely in the past to de- 
scribe many types of modified stem found in the genus. In this account, “rhi- 
zome” is reserved for those stems which are dorsiventral, creeping, main 
axes, prostrate on or under the ground, progressively producing aerial 
shoots in a sympodial pattern. Such rhizomes, however, can be separated 
into three groups. 

i. Slender elongated rhizomes producing aerial shoots every other 
node. These are clothed in leaflike scales which may or may not over- 
lap. Examples of this form of rhizome are to be found in K. elatior, 
K. brevifolia and K. intricata [Fig. 3 (a)]. 

ii. Slender or stout compact rhizomes bearing aerial shoots at each 


Studies in Cyperaceae in southern Africa: 10 


265 


No of 

individuals 



Wing width in millimetres 

Fig. 2 

The range of wing width distribution in twenty individuals of K. alba (solid line) and 

K. alata (dotted line). 


node. Examples of this type are to be found in K. erecta, K. melanos- 
perma and K. polyphylla. [Fig. 3 (b)]. 
iii. Uniseriate rows of woody stem-bases or tufts of woody stem-bases 
clothed in fibrous sheaths as are found in K. alba, K. alata and K. 
odorata, all three of which are associated with grasslands. [Fig. 3 (c)]. 
In contrast to these, three species possess stolons. These lateral branches 
arise at or below the surface and radiate from the mother plant producing 
aerial shoots some distance from the mother plant. K. albiceps (Ridley) 
Rendle and K. pulchella have extremely slender stolons which are frequent- 
ly damaged unless collected with great care. Those of K. nemoralis are more 
robust [Fig. 3 (d)j. 


266 


Journal of South African Botany 



Fig. 3 

Underground portions of Kyllinga species: a. Elongated rhizomes of K. elatior. b. 
Congested rhizomes of K. polyphylla. c. Stem-bases of K. alba. d. Stolons of K. 
nemoralis. Scale in each case is equivalent to 10 mm. 

The problems of imposing a single-character classification upon a vari- 
able genus such as this are evident when considering the subgenera of Kyl- 
linga. 

Choice of a wing on the keel of the glume to distinguish subgenus Alata 
cannot be upheld in view of the range of variation found. The use of number 
of nuts to define subgenus Pseudo-Pycreus is difficult to justify when some 
species have one or two nuts. In this account therefore, subgenera are not re- 
cognised. If any characteristic were chosen to distinguish groups within this 
genus, it would be the underground stem, but even here it would lead to ar- 
tificial groups. 

At the species level, however, the nature of the underground stem has 
been very useful, together with inflorescence form including shape of head, 
bract and colour. 


Studies in Cyperaceae in southern Africa: 10 


267 


Kyllinga Rottb0ll nomen conservanda Kyllinga Rottb., Descr. Ic. Nov. PI. 
12 (1773) p.p.; Vahl, Enum. PI. II 379 (1806); Nees in Linnaea 9: 286 
(1834); Kunth, Enum. PI. II: 127 (1837); Boeck. in Linnaea 35: 403 (1868) 
(excl. sect. Pseudo-Kyllingia)\ Benth. and Hooker, Gen. PI. 3: 1045 (1883); 
C. B. Clarke in Hooker, PI. Brit. India 4: 587 (1893), in Thiselton-Dyer, 
FI. Cap. 7: 151 (1897) and in Thiselton-Dyer, FI. Trop. Afr. 8: 266 (1901); 
Schonland, Intr. to South Afr. Cyper. 19 (1922); Brain in Pro. Rhod. Sc. 
Assoc. 33: 78 (1934); Hutchinson, in Hutchinson and Dalziel, FI. West 
Trop. Afr. 2, 2: 486 (1936); Phillips, Gen. South Afr. FI. PI. 153 (1951). 
Thryocephalon J. R. and G. Forster, Char. Cen. PI. 129, t 65 (1776). 
Killingia Juss., Gen. PI. 27 (1789). 

Kylingia Stokes in Bot. Mater. Med. 1: 120 (1812). 

Kylinga Roem et Schult., Syst 2: 6 and 236 (1817). 

Kyllinga Lestib., Ess. Cyper. 28 (1819); Blanco, PI. Filip. 23 (1845) — 
Killyngia Hamilt., Prodr. PI. Ind. occ. 3 (1825). 

Hedychloe Raf. in Ann. Nat. 16(1820). 

Cyprolepis Steud. in Flora 33: 229 (1850). 

Lyprolepis Steud., Syn. Cyp. 130, (1855). 

Cyperus subgenus Kyllinga Pax in Engl, and Prantl. Nat. pflanzenfam. 2, 
2: 98 (1887); Kiikenthai in Engler, Pflanzenr. 4, 20: 556-614 (1936); Tour- 
nay in Robyns, Flore des Sperm, du Parc. Nat. Albert. Ill: 223 (1953). 

Type: K. monocephala Rottb. nom. illeg. 

(K. nemoralis (J. R. & G. Forst.) Dandy ex Hutchinson and Dal- 
ziel) . 


Key to Species 

1. Inflorescence dense, of many spikelets (always exceeding 20), occasionally ac- 
companied by vegetative shoot (pseudovivipary) • ^ 

Inflorescence lax, of 15-20 spikelets, or fewer, some replaced by a vegetative 

K. pauciflora Ridley 

2. Plants with stolons terminating in aerial stems: in herbarium material these 
may appear simply as individual plants unless the underground portion is ^ 

Plants (hizomatous, series of aerial stems produced on elongated or con- 
gested rhizomes sometimes with fibrous sheaths; or tufted ........ • • • • • ■ ■■■ ■ - 

3. Inflorescence of 3-1 dark red, sessile, cylindrical spikes . . . . . k. puichella kunth 
Inflorescence straw-coloured or green, never red; usually spheroidal . . • ■ • • • • 

4. Stolons fleshy, white, radiating in all directions from "lain plant, s P lke lf s 

winged, maturing one nutlet K. nemoralis J. R. 5 & ■*«*•) Dandy 

Stolons slender, not fleshy, covered with imbricate scales, usually 1 stolon per 
main plant; spikelet wingless, maturing two 

5. Plants tufted, or with a uniseriate row of woody, swollen stem-bases clothed ^ 

in fibrous sheaths 


268 


Journal of South African Botany 


Plants rhizomatous; rhizome either congested or elongated; stem-bases not 
clothed in fibrous sheaths 9 

6. Inflorescence cylindrical; stamens 2 or 1 per floret 7 

Inflorescence globose; stamens 3 8 

7. Inflorescence pale green to buff when dry; central spike 4-7 mm long; keel of 
glume ciliate to slightly winged in older spikelets; anthers 1 or 2 per floret; 

nut narrowly ellipsoid, golden K. welwitschii Ridley 

Inflorescence white; central spike 8-16 mm long; keel of glume wingless, gla- 
brous or with 1 or 2 hairs; anthers 2 per floret; nut broadly obovoid, yellow 
when young turning red-brown later black K. odorata Vahl. 

8. Inflorescence golden-green, spikelets usually markedly winged . . . K. alata Nees 

Inflorescence white or cinnamon K. alba Nees 

9. Rhizome slender, aerial stems distant from one another, produced every 

other node 10 

Rhizomes compact, aerial stems contiguous, produced at every node 12 

10. Stamens 1 per floret; rhizome slender, white with pale brown scales shorter 

than internodes K. brevifolia Rottb. 

Stamens 3 per floret rhizomes not as above 11 

11. Inflorescence with 1-3 spikes, central spike elliptic or cylindrical; bracts long 
(up to 290 mm), horizontally orientated; rhizome 2-3 mm in diam. with 

transparent membranous scales with striate markings K. elatior Kunth 

Inflorescence 1 globose to ovate head: bracts 3 short, lowest erect and later 
reflexed; rhizome 1-2 mm in diameter with loose, red-brown, overlapping 
scales K. intricata Cherm. 

12. Plants generally less than 400 mm; rhizome a row of bulbous-based, contigu- 

ous stems clothed in scales (when dry, stem-bases shrunken within the 
scales), inflorescence a single, globose, golden or green spike . K. erecta Schum. 
Plants generally taller, 300-640 mm; rhizome compact, stout (5-8 mm in di- 
ameter) not shrinking conspicuously when dried; inflorescence either cylindri- 
cal or a group of ovoid or globose heads 13 

13. Inflorescence 1, usually cylindric, yellow-green, but subglobose and rust 
streaked under unfavourable conditions, subtended by 3 stiff erect bracts re- 
flexing to 45 ° with stem later, with possible 4th very short (120 mm long) 

bract K. melanosperma Nees 

Inflorescence invariably of more than 1, sessile, globose, green spikes sub- 
tended by 3-8 spreading bracts, 130-210 mm long 

K. polvphylla Wild, ex Kunth 


1. K. pauciflora Ridley in Trans. Linn. Soc Ser. 2, 2: 147, t.23, fig 1-4 
(1884); C.B. Clarke in Durand and Schinz, Conspect. FI. Afr. 5:530 (1895), 
in Thistleton-Dyer, FI. Cap. 7:152 (1897) and in Flora Trop. Afr. 8:273 
(1902), Rendle, Catal. Afr. PI. Welw. 2:105 (1899); Medley Wood in Trans. 
S.A. Phil. Soc. 18, 2:250 (1908). Type: Angola, Welwitsch 6811 (BM, 
holo.). 

Cyperus ridleyi Mattf. and Kiikenth. in Engler, Pflanzenr. 20:599 (1936). 

Rhizome slender, creeping, elongated, or stems congested and tufted. 
Culms slender, 80-600 mm tall, base scarcely thickened, covered with few 
sheaths. Leaves few, shorter than culm, blades narrow (1-3 mm wide), acu- 
minate. Bracts 3, short (30-100 mm) lowest upright, others spreading. 


269 


Studies in Cyperaceae in southern Africa: 10 

Inflorescence of 1 sub-lateral spike, lax, rarely more than 15 spikelets, 
golden-green, often exhibiting adventitious plantlets. Spikelets long for the ge- 
nus. 3,4—6,54 mm, ovate-lanceolate in outline, fertile glumes 2, unequal, well de- 
veloped, plus third upper reduced glume; lower glume 7-9-nerved, subtend- 
ing bisexual floret, middle 5-7-nerved subtending male (rarely bisexual 
floret), upper empty, keels green, glabrous or slightly spinulose, excurrent 
into pointed slightly recurved mucro (0,20 to 0,72 mm). Stamens 2, anther 
connective slightly produced. Style long but often exceeded in length by 
style branches. Nut 1, half length of glume, obovoid, golden-brown. 

TRANSVAAL— 2330 (Tzaneen): Rosendal Dam, Letaba District (-CD) 
Scheepers 1190 (PRE). 

—2730 (Vryheid): Wakkestroom (-AC), Devenish 952 (PRE). 

NATAL— 2829 (Harrismith): Cathedral Peak Forest Reserve (-CC), Killick 
1103 , 1273 (NH, NU). 

—2929 (Underberg): Estcourt district, Hlatikulu, Giants Castle area (-BA), Edwards 
2242 (PRE); Ntabamthlope vlei, Estcourt (-BC), Downing 124, 142, 153, 156 (NU), 
Getliffe 298 (NU); Mpendhle, Tillietudlem (-DB), Huntley 443 (NU). 

— 2930 (Pietermaritzburg): Lions River district, roadside near Rawdon’s Hotel 
(•AC), Gordon-Gray s.n. (NU), Getliffe 224 (NU); Hermansberg district, vlei near 
school (-BB), Getliffe 174 (NU). 

K. pauciflora was described as “having the habit of one of the few 
flowered Marisci while the spikelets suggest those of K. aurata Nees” Ridley 
(1884). It was distinguished on the basis of its longer spikelets, few in num- 
ber, and the presence of “vivipary”. (Fig 4b.) 

The identity of several Natal gatherings with lax golden inflorescences of 
very few spikelets (Huntley 443, Killick 1273 and Willd 1439), a gathering 
from Zimbabwe, has been subject to some debate. Kew determinations 
placed them in K. pauciflora but Robinson (1959) associated them with K. 
intricata. Robinson also suggested that some of his gatherings deserved 
specific rank as a new species (sp. viii). 

Examinations of the Natal specimens and of populations in the field 
showed that they could not satisfactorily be placed with K. intricata as they 
possessed a more congested rhizome and inflorescences of a maximum of 20 
spikelets. The phenomenon called vivipary by Ridley (loc. cit.) was well rep- 
resented in the Natal populations. Comparison with Robinson’s proposed 
“species viii” suggested these were very similar. 

The production of adventitious plantlets is a conspicuous feature of these 
populations and may be associated with the end of the growing season or 
with waterlogged conditions. Rhizome structure is interesting and may also 
exhibit seasonal variation. Elongate internodes of older rhizomes terminate 
in a cluster of closely packed culms. Herbarium specimens may include both 
portions if adequately collected or just the terminal congested young por- 


270 


Journal of South African Botany 



A. Morphology of the rhizome of K. pauciflora with congested stems in young por- 
tion and elongated older portions. B. Vegetative shoot replacing one of the few 
spikelets in the head. Scales in millimetres. 


tions. The unique combination of vegetative reproduction and few flowers 
suggests that all these belong to K. pauciflora. 

Ridley (1884) commented on the longer spikelets of K. pauciflora and a 
comparison of our material with tropical K. pauciflora showed a tendency 
for our material to have slightly smaller spikelets. There is however no clear 
distinction between the tropical material averaging 5,09 mm, Robinson’s 
“sp. viii” averaging 4,8 mm and the Natal gatherings averaging 4,13 mm but 
rather a gradual decrease in size towards the southern limits. 

No detailed studies of the breeding strategies of these populations have 
been carried out to date and the correlation of vegetative propagation with 
waterlogged marshy conditions has been observed frequently with this and 
other species, but not tested. Nor can one entirely rule out the possibility 
that the Natal populations represent hybrid swarms possibly derived from 
K. erecta Schum. Until these questions are resolved, it is proposed that the 
material be retained in K. pauciflora Ridley. 


Studies in Cyperaceae in southern Africa: 10 


271 


2. K. pulchella Kunth, Enum. PI. 2: 137 (1837); Boeck in Linnaea 35: 405 
(1868); C. B. Glarke in Durand and Schinz, Conspect. FI. Afr. 5: 531 
(1895), in Thiselton-Dyer, FI. Cap. 7: 154 (1897), in Thiselton-Dyer, Flora 
Trop. Afr. 8: 284 (1901) and in Kew Bull. Addit. Ser. 8: 94 (1908); Burtt 
Davy and Pott-Leendertz in Ann. Tvl. Mus. 132 (1912); Phillips in Ann. S. 
A. Mus. 16: 319 (1917); Podlech, in Merxm. et al., Prodr. FI. Siidwestafr. 
165: 31 (1970). Type: Cape, Albert Division, Drege 7384 (K!, holo.). 

K. atrosanquinea Steud. in Flora 598 (1842) and Syn. PI. 2: 69 (1855). 
Type: Ethiopia, Schimper 1269. 

Cyperus teneristolon Ktikenth., in Engler, Pflanzenr. 4, 20: 574 (1936). 

Stolons very slender (less than 1 mm in diam.). Stem-bases slightly thick- 
ened. Culms slender, up to 430 mm tall, triangular, solitary or in small 
groups, glabrous, glaucous-green, leafy at base. Leaves often as long as 
culms, glabrous, flaccid. Bracts 3-4 (see footnote), leaflike, lowest erect, up 
to 120 mm long, others shorter spreading. Inflorescence 1-3 spikes, sublate- 
ral when immature, middle spike ovate, up to 16 mm long and 10 mm wide, 
dark red. Spikelets many, 3-5 mm long, spreading, 3-5 fertile glumes, 2-4 
bisexual florets, 3-1 nuts. Fertile glumes ovate in outline, lowest 9-nerved, 
middle 7-nerved, upper 3-5-nerved; keel green or greenish-yellow, wingless, 
glabrous or with few hairs, excurrent into short mucro. Stamens 3, anthers 
linear, long, connective produced. Style long, exserted, stigmas 2. Nut ellip- 
soid to obovoid, half length of glume, chestnut brown. 

TRANSVAAL — 2628 (Johannesburg): Frankenwald (-AA), Gilliland 25055 
(PRE); Boksburg (-AB), Munay s.n. (PRE 29315). 

— 2729 (Volksrust): 29 km from Volksrust on Amersfoort Road (-BD), Hilliard 2991 
(NU). 

Without precise locality: Schlechter 5529 (PRE). 

ORANGE FREE STATE— 2926 (Bloemfontein): Reddersburg district (-CA), 
Acocks 18696 (PRE). 

NATAL— 2828 (Bethlehem): Royal Natal National Park (-DB), Getliffe 22 
(NU). 

— 2929 (Underberg): Mpendhle, 20 km from Dargle (-DB), Moll 680 (NU). 

— 2930 (Pietermaritzburg): Lions River district, Karkloof (-AC), Gordon-Gray s.n. 
(NU); Mount Gilboa, Karkloof, Getliffe 2058 (NU). 

LESOTHO — 2929 (Underberg): Sehlabathebe Reserve (-C C),J acot Guillarmod, 
Mzamane and Getliffe 91 (PRE, INCOL). 

CAPE— 3025 (Colesberg): 16 km SW Colesberg (-CA), Acocks 16310 (PRE). 

— 2337 (Stutterheim) Mount Cume, Dohne (-CB), Acocks 9495 (PRE); Hilner ~59 
(PRE); Komgha (-DB), Flanagan 1261 (K, NH, PRE). 

Also found in Ethiopia, Kenya and Tanzania. ___ 

Acocks 18696 consisted of two plants one of which has about 14 bracts. Each of the 
four heads comprising the inflorescence appears to have a full complement ot bracts. 
This is, however, an exception to the general rule that bract number in the species is 
3 or 4. 


272 


Journal of South African Botany 


Kyllinga pulchella is distinguished from all other southern African 
species by its dark red inflorescences. It has not been widely collected 
though Clarke (1901) remarked that it was “frequent in South Africa”. 
Careful collection is necessary to include the characteristic delicate stolons. 
A tropical species of K. anomala Peter & Kukenth. may be its nearest rela- 
tive, but is distinguished from K. pulchella by its peduncled axillary spikes 
and red keels to the glumes whereas those in K. pulchella are green. These 
distinctions may not be sound and Napper (1971) has amalgamated both 
into K. pulchella. K. pulchella is a distinct species in South Africa differing 
from the only other species with slender stolons, K. albiceps , in the striking 
red colour of its spikes. 

3. K. nemoralis (J.R. and G. Forster) Dandy, ex Hutchinson and Dal- 
ziel. Flora W. Trop. Afr. 2, 2: 487 (1936), Nelmes and Baldwin in Amer. 
Journ. Bot. 39 : 389 (1952), Hooper in Hepper, Flora W. Trop. Afr. ed. 2, 
3 : 307 (1972) 

Thryocephalon nemorale J. R. and G. Forster, Charact. Gen. PI. 129, 
t 65 (1776). 

Kyllinga monocephala Rottb., Desc and Icon Nov. PI. 13, fig 4 (1773); 
Nees, in Wight, Contrib. Bot. of India 91 (1834); Kunth, Enum. PI. 2: 129 
(1837); Steud., Synops, Cyp. 2: 67 (1855); Boeck. Linnaea 35: 427 (1868) 
p.p, Ridl. in Trans. Linn. Soc. 2. Ser Bot 2: 147 (1884); C. B. Clarke, in 
Hooker, FI. Brit. Ind. 6: 588 (1893), in Durand and Schinz, Conspect. FI. 
Afr. 8: 530 (1895), in Thiselton-Dyer, FI. Trop. Afr. 7: 272 (1902) and II- 
lustr. Cyp. t ii. f 1-2 (1909); Chermeson, Catal. PI. Madag. 8 (1931). Type: 
Rottb., Desc. et Icon. Nov. PI. 13, fig. 4. (Iconotype). 

Cyperus kyllinga (Endl.) Kukenth. , in Engler Pflanzer. 4, 20:606 (1936). 

Plants with stolons, 10-100 mm long, slender, white and pliable when 
fresh, fragrant, about 2 mm in diam., with membranous, pink-brown, striate 
scales, developing fine feathery roots. Culms usually distant, occasionally 
clustered, 40-570 mm tall, flaccid (except Getliffe 193), triangular, ba r -s not 
thickened. Leaves usually shorter than culms, sheaths membranou west 
pinkish-brown, striate; blades 3-5 mm wide, flat, margins and keel brid 
particularly near apex, flaccid (except Getliffe 193). Bracts 3-4, 1 100-1 300 
mm long, clasping inflorescences, upright later reflexed or spreading. Spike- 
lets many, ovate elliptic in outline, 2,95-5 mm long, floret 1 bisexua rarely 
second male floret, fertile glumes 2 sometimes with third reduced hyaline 
glume, sides of both well-developed glumes obscurely nerved, 3-4 promi- 
nent nerves close to thick fleshy green winged keel that is excurrent into 
slightly recurved mucro 0,5-0, 8 mm long; wing 0,1-0,15 mm wide, spinu- 
lose. Stamens 3. Style long, slender, bifid. Nut 1, less than \ length of 


Studies in Cyperaceae in southern Africa: 10 


273 



Morphology and distribution of K. pulchella in southern Africa. A. Basal portion of a 
shoot with slender stolons terminating in erect shoots. B. Multiple heads often toun 
in this species. C. Cross-sectional outline of the culm. D. Dark red spikelet ( ower 
empty glumes, a bract and prophyll, removed) with three fertile glumes maturing a 
least two nutlets. E. Distribution of K. pulchella in southern Africa. Figure sea es in 
millimetres: map scale in kilometres. 


274 Journal of South African Botany 

spikelet, obovoid to oblong, butter yellow turning red-brown, later black, 
slightly apiculate. 

NATAL — 2832 (Mtubatuba): 16 km N of Mtubatuba (-AC), Getliffe 59, 123 
(NU); Dukuduku (-AD), Ward s.n. (NU). 

— 2930 (Pietermaritzburg): Isipingo Beach (-DD), Ward 1250 (NU). 

— 2931 (Stanger): Durban Country Club Golf Course (-CC), Getliffe 193 (NU); Dur- 
ban Botanic Garden, Getliffe 996 (J). 

K. nemoralis has been described as “very general in warmer parts of the 
Old World particularly near the sea” (Clarke, 1897) and as a weed of sec- 
ondary forest and disturbed areas (Hooper, 1972). Clarke (1897) cited only 
two doubtful records from Africa but it has subsequently been included in 
accounts of African material (Kukenthal, 1936; Napper, 1971; Hooper, 
1972b) with the most southern record in Mozambique. During this investiga- 
tion five records of this species have been studied. All were coastal, two in 
areas which have been subject to considerable disturbance (a golf course 
and a densely populated resort) and a third on the fringe of a pond in the 
Botanic Garden where it is frequently mowed. 

These Natal gatherings agree with descriptions of K. nemoralis though 
the term stolon is preferred here to “slender rhizome” (Napper, 1971). One 
distinctive feature not previously recorded for this species, is the presence of 
“vivipary” in these populations. Small adventitious propagules arise in the 
inflorescence, apparently replacing a spikelet. In Cyperus prolifer Lam., this 
phenomenon was observed to be correlated with disturbance and waterlog- 
ging (Getliffe & Baijnath, 1976). In Kyllinga nemoralis it may reflect a reac- 
tion to similar conditions. Once again, the underground portions are the key 
diagnostic character and this is nof always included in herbarium specimens. 
Without reference to these slender stolons the species may be confused with 
K. brevifolia or K. elatior , but it differs from both in the presence of a winged 
keel to the glumes. Its fleshy stolons distinguish it from the other stolonifer- 
ous species, K. pulchella and K. albiceps, in which the stolons are very slen- 
der. 

4. K. albiceps (Ridley) Rendle in Cat. Afr. PI. Welw. 2: 106 (1896), 
C. B. Clarke in Thiselton-Dyer, Flora Trop. Afr. 8: 286 (1901); Hutchinson 
in Flora W. Trop. Afr. 2, 2: 486 (1936). 

K. macrocephala A. Rich var. angustior C. B. Clarke, Conspect. FI. Afr. 
5: 529 (1895). Type: Scott Elliott 7626 (K, holo.!) 

K. merxmuelleri Podlech, Mitt. Staatssamml. Munch. 3: 525 (1960), in 
Prod. FI. Siidwestafr. Fam. 165. L. 16:31 (1967). Type: Vley 16 ml Ostlich 
Runtu. Merxmueller and Giess 2136 (PRE, iso.!). 


Studies in Cyperaceae in southern Africa: 10 275 



Morphology and distribution of K. nemoralis. A. Inflorescence with lateral subsidi- 
ary spike. B. Cross-sectional outline of culm. C. Single winged spikelet (lower empty 
bract and prophyll removed). D. Portion of shoot with prominent stolons radiating 
from parent plant. E. Distribution along Natal coast. Figure scales in millimetres: 
map scale in kilometres. 


276 


Journal of South African Botany 


Cyperus albiceps Ridley, Journ. Bot. 22: 16 (1884). Type: Congo, Chris- 
tian Smith (BM, holo.). 

Cyperus richardii Steud. var. angustior (C.B. Clarke) Kukenth. in Eng- 
ler, Pflanzenr. 4, 20: 570 (1936). 

Stem-bases woody, swollen, sheathed, stoloniferous. Stolons 1 mm wide, 
up to 500 mm long, covered with bracts which later split into fibres. Culms 
solitary, up to 300 mm long, slender, compressed triangular, glabrous, leafy 
at base. Leaves shorter than culms, 2-5 mm wide, flaccid, acuminate. Bracts 
2-3, erect, later reflexed, leaflike, longest up to 95 mm long. Inflorescence 
of 1 globose or ovoid spike or 2-3 spikes confluent into subglobose head, up 
to 8 mm in diam, straw-coloured. Spikelets many, ovate-elliptical in outline, 
less than 3 mm long, 2-flowered, 2-3 fertile glumes. Fertile glumes straw- 
coloured, sometimes gland-dotted, 7-9-nerved, keel probably green when 
fresh, fading to straw, excurrent into short mucro. Stamens 3. Style almost 
equal in length to style branches. Nuts 2, ellipsoid to oblong, half length of 
glume, golden-brown. 

S.W.A. — 1719 (Runtu): Vley 16 ml Ostlich Runtu (-DD), Merxmueller and 
Giess 2136 (PRE). 

— 1814/1816: Ovamboland, de Winter and Giess 6900 (PRE). 

K. macrocephala var. angustior was differentiated from K. macrocephala 
by its smaller inflorescence and shorter spikelets but transferred by Clarke 
(1901) to K. albiceps (Ridley) Rendle with the remark that it differed very 
little from K. macrocephala , but for the slender stems. There are undoubt- 
edly two entities here; a tropical group and one which also has a southern 
element. Both were examined and the difference in spikelet length and 
inflorescence length was confirmed. A more significant difference however is 
the difference in the number of bisexual florets. In K. macrocephala , more 
than two nutlets were produced while the group assigned to the variety of K. 
albiceps , have only two fertile florets. The type of Clarke's variety, K. 
macrocephala var. angustior (Scott Elliott 7626) bears the comment in 
Clarke’s hand that it has three nuts. However, examination of this specimen 
could not confirm this and the material examined suggests that two nutlets 
are characteristic of the variety. The isotype of K. merxmuelleri and a sec- 
ond gathering cited by Podlech (1960) were examined and found to agree in 
every respect with K. albiceps which epithet has precedence. 

This species shares the delicately stoloniferous habit of K. pulchella but 
its buff inflorescences are readily distinguished from the dark red inflores- 
cences of K. pulchella. 

5. K. welwitschii Ridley, Trans. Linn. Soc. Ser. 2, 2: 147 (1884); Hooper 
in Hepper, Flora W. Trop. Afr. ed 2, 3: 305 (1972). Type: Angola, Pungo 
Andongo, Welwitsch 6796 (K, holo.; BM, iso.). 


Studies in Cyperaceae in southern Africa: 10 


277 


K. blepharinota Hochst. ex Boeck., in Linnaea 35: 414 (1868). 

A:, caespitosa Ridl. (non Nees) in Trans. Linn. Soc. Ser. 2, 2: 145 (1884). 

K. triceps var. ciliata Boeck. in Linnaea 35: 414 (1868); C. B. Clarke in 
Durand and Schinz, Conspect. FI. Afr. 5: 533 (1895) and in Thiselton-Dyer, 
Flora Trop. Afr. 8: 281 (1901); Kiikenthal in Engler, Pflanzenr. 4, 20: 579 
(1936). Type: Abyssinia, Schimper 2201 . 

K. controversa var. sub exalata C. B. Clarke in Thiselton-Dyer, Flora 
Trop. Afr. 9: 271 (1901). 

Cyperus controversus var. subexalatus (C. B. Clarke) Kiikenth., in Eng- 
ler, Pflanzenr. 4, 20: 612 (1936). 

Plants tufted or with short row of thickened, woody, stem-bases clothed 
in persistent, tough sheaths. Culms 40-50 mm tall, very slender, triangular, 
furrowed, glabrous. Leaves shorter than culms, blades 1-2 mm wide, gla- 
brous or margins and keel minutely scabrid, flaccid, markedly acuminate. 
Inflorescence of 3-1 spikes, pale green to buff-coloured when dry, dense; 
central oval, 4-7 mm long, lateral smaller, subglobose. Bracts 3, up to 8 mm 
long, sometimes with 1 additional, very reduced bract. Spikelets numerous, 
elliptic in outline, 1,5-2, 5 mm long; fertile glumes 3; lowest subtending bi- 
sexual floret, white scarious gland-dotted, obscurely 5-nerved, keel probably 
green, scarcely excurrent into mucro, ciliate, slightly winged in older spike- 
lets; middle glume fully developed, empty, otherwise as in lowest glume but 
3-nerved; uppermost glume minute, hyaline. Stamens 2-1. Style and style 
branches shorter than mature nut, style branches exceed style in length. Nut 
1, narrowly ellipsoid, approximately 1,5 x 0,5 mm, golden. 

S.W.A. — 2115 (Karibib): Okongava, Granitbank (-BB?), Dinter 6961 (PRE); 
Farm Neuschwaben (-DD), Kinge 3048 (PRE). 

Also found in West Africa, Tanzania and Mozambique. 

The taxonomic delimination of K. welwitschii Ridley has had a varied 
history. Clarke (1895) cited, under K. triceps var. ciliata , nineteen African 
specimens including eight Welwitsch gatherings which, according to Robin- 
son (1959), are plants with immature spikelets. One of these is the type of 
K. welwitschii Ridley. In 1901, Clarke reorganised K. triceps var. ciliata 
leaving the Welwitsch gatherings in that variety but transferring the remaining 
specimens to K. controversa var. subexalata C. B. Clarke. The only differ- 
ence between these two groups of specimens was the presence, in the latter, 
of a very narrow wing on the keel of the glume. No measurements were 
given but the varietal epithet suggests it was a very narrow wing. The Wel- 
witsch gatherings remaining in K. triceps var. ciliata were immature and this 
is significant for in the present study it was found that wing width was a fac- 
tor of the age of the spikelets. 

BJ — 7 


278 


Journal of South African Botany 


Thus, the two varieties cannot be upheld but must be regarded as one 
taxon. Present studies show that the difference between this taxon and K. 
triceps are real and of sufficient significance to warrant specific recognition, 
however it cannot be assigned to K. controversa as they are quite distinct 
from this rather doubtful species. K. welwitschii Ridley is the oldest valid 
name and is consequently applied to this taxon. This species is distinguished 
from other tufted species by its pale green multiple heads of sessile spikes, 
and the prescence of 2 (or 1) stamens. 




w Fig. 7 

A. Spikelet of K. welwitschii with two fertile winged glumes (lower prophyll and 
bract removed). B. Slender elongated nutlet removed from spikelet. C. Plant with 
slender stolon terminating in a new shoot. All scales in millimetres. 

6. K. odorata Vahl, Enum. PI 2: 382 (1806); Nees, Linnaea 9: 286 (1834); 
Kunth, Enum. PI. 2: 132 (1837); Boeck., Linnaea 35: 410 (1868); C. B. 
Clarke, Kew Bull. Add. Ser. 8: 93 (1908); Hutchinson in Hutchinson and 
Dalziel, Flora W. Trop. Afr. 11 , 2: 487 (1936); Hooper in Hepper, Flora W. 
Trop. Afr. 2: 304 (1972). Type: Habitat in Americo meridional v. Rohr. 
Richard (not traced) 

Cyperus sesquiflorus (Torr.) Mattf. and Kiikenth. in Engler, Pflanzenr. 
4,20: 591 (1936). 


Studies in Cyperaceae in southern Africa: 10 279 

C. sesquiflorus var. cylindricus (Nees) Kukenth. in Engler Pflanzenr 4 
20: 592 (1936). 

K. cylindrica Nees in Wight, Contrib. Bot. India: 91 (1834) and Linnaea 
9: 286 (1934); Kunth, Enum. PI. 2: 133 (1837); Boeck., Linnaea 35: 415 
(1868); C. B. Clarke in Hooker f., FI. Brit. India 6: 588 (1893); in Durand 
and Schinz, Conspect. FI. Afr. 5: 528 (1895), in Thiselton-Dyer, Flora Cap. 
7: 153 (1897), in Thiselton-Dyer, Flora Trop. Afr. 8: 282 (1901) and in Kew 
Bull. Add. Ser. 8: 93 (1908); Medley Wood, in Trans. S.A. Phil. Soc. 17, 2: 
250 (1908). Type: India, herb. Royle no. 39 (LIV, holo). 

K. triceps Rottb. var. obtusiflora Boeck., Linnaea 414 (1868). 

Cyperus triceps var. obtusiflorus (Boeck.) Kukenth. , in Engler, Pflan- 
zenr. 4, 20: 579 (1936). 


Roots fibrous. Plants tufted or forming short row of old sheathed stem- 
bases. Culms up to 540 mm tall, triangular, ridged and furrowed, glabrous, 
scarcely thickened at base. Leaves many, tufted at base of culm, half length 
of culms; sheaths short, brown; blade 2-4 mm wide, margins and keels scab- 
rid. Inflorescence of 1-3 dense, white spikes; central cylindric, 8-16 mm 
long, 5-7 mm wide; lateral small, globose to cylindric. Bracts 3-4, leaflike, 
longest up to 93 mm long. Spikelets many, 2,47-2,52 mm long, broadly 
ovate in outline, tapering to acute apex, 1- (rarely 2)-flowered, fertile 
glumes 2, almost equal, lower 9-nerved, upper 5-7-nerved, sometimes en- 
closing third reduced hyaline glume; keel wingless, smooth or with 1-2 
hairs, green, keel of lower glume slightly excurrent into mucro. Anthers 2. 
Style short, exceeded in length by style branches. Nut broadly obovoid, apex 
truncate, § length of spikelet, yellow to red-brown, later black. 

TRANSVAAL— 2330 (Tzaneen): Westfalia Estate (-CA), Scheepers 17 (PRE). 

- — 2430 (Pilgrims Rest): Mariepskop district (-DB), van der Schijff 4946 (PRE). 

NATAL— 2830 (Dundee): 3 miles N of Kranskop (-DC), Getliffe212 (NU). 

— 2831 (Nkandhla): Shongweni (-AA), Ross 773 (NU); Ngoya Reserve 

(-DC), Getliffe 216 (NU); Mtunzini “Hamewith” (-DD), Mogg 4428 , 4522 (PRE). 

— 2930 (Pietermaritzburg): Burdons Farm, Karkloof (-AC), Wells 1089 (NU); “The 
Dargle”, Getliffe 179 (NU); 3 miles up Curry’s Post Road from Balgowan, Downing 
s.n. (NU); “The Start”, Getliffe 203 (UN); “Rosebank”, Durban Road, Greytown 
(-BA), Getliffe 195 (NU); Ahren, Mowbray, Fisher 953 (NU, PRE); Hermannsburg, 
vlei near school (-BB), Getliffe 172{ NU); Noodsberg Road, past Montebello turnoff 
(-BD), Getliffe 1658 (NU); Kings Cliff, Noodsberg Road, Getliffe 166 (NU); Town 
Bush Valley (-CB), Ward 117b (NU), Ross 721 (NU); Chase Valley, Getliffe 24 
(NU), Ross 729 (NU); Portals Farm, Raisethorpe, Ram s.n. (NU); Pietermaritzburg 
Country Club, Ross 726 (NU); Thomas Moore School, Fields Hill, Getliffe 175 
(NU); Pinetown Govt. Indian School, Getliffe 229 (NU). 

— 2931 (Stanger): Howard College (-CC), Getliffe 177 (NU). 

CAPE— 3129 (Port St. Johns): Port St. Johns above Tiger Flats (-DA), Galptn 
2846 (PRE). 


280 


Journal of South African Botany 


K. odorata is based on American material and K. cylindrica on Indian 
specimens. African gatherings have been assigned to both of these, but 
Kiikenthal (1936) regarded them as two variations of C. sesquiflorus. 

K. odorata was relegated to C. sesquiflorus var. sesquiflorus distin- 
guished by its wider inflorescence and longer spikelets from the variety 
cylindricus (= K. cylindrica Nees). 

Particular attention was therefore paid to these two characters in a sur- 
vey of African material, including Welwitsch gatherings cited by Kiikenthal. 
All specimens seen had inflorescences well within the range of the type and 
the spikelet length ranged from 2,3 to 3,25 mm without any discontinuity 
into two groups. Doubts about the separation of the taxon into two varieties 
were confirmed when specimens seen by Kiikenthal did not fit his delimita- 
tions. It is apparent that no purpose is served in splitting this variable ma- 
terial into two varieties; accordingly only K. odorata Vahl is upheld. This 
species is characteristic of grassland habitats and its distinctive cylindrical 
head, frequently subtended by two smaller globose heads makes it readily 
distinguishable from other grassland forms. The nuts ripen to glossy black 
and are visible through the glumes giving mature heads a mottled appear- 
ance. 

7. K. alata Nees, Linnaea 9: 281 (1834), Linnaea 10: 139 (1835-36): Kunth, 
Enum. PI. 2: 136 (1837); Steud., Synops. Cyp. 69 (1855); Boeck, Linnaea 
35: 430 (1868). Type: locis humidis arenosis ad “Koegakammaklog” et 
flumen Zwartkopsrivier alt II, Ecklon 883. (S, iso.?!). 

Kyllinga alba var. alata (Nees) C. B. Clarke in Durand and Schinz, 
Consp. FI. Afr. 5: 526 (1895), in Thiselton-Dyer, Flora Trop. Afr. 8:272 
(1901); Burtt Davy and Pott-Leendertz, Ann. Trans. Mus. 132 (1912); Phil- 
lips, Ann. S.A. Mus. 16: 318 (1917); Schonland, Mem. bot. Surv. S. Afr. 3: 
19, t. 1 (1922). 

Cyperus alatus (Nees) F. Muell., Fragm. 8: 272 (1874); Kukenth. in 
Engler. Pflanzenr. 4, 20: 611 (1936). 

Plants tufted, forming a short row of sheathed stem-bases, sheaths 
brown, tough, fibrous. Culms frequently, but not invariably, ciliate-pubes- 
cent below inflorescence. Leaves numerous, flaccid, generally more than 
half length of culm; sheaths less than 3 length of culm, blade up to 5 mm 
wide, margin and keel scabrid, especially in young leaves. Inflorescence a 
single ovate to globose spike, rarely 2 additional lateral peduncled spikes 
[ Getliffe 305 (NU),], golden-green. Bracts 3-4, occasionally reduced to 2. 
Spikelets many, approximately 6 mm long, ovate-lanceolate in outline, fer- 
tile glumes 2, almost equal, subtending 2 bisexual florets or upper male, 
both keels green, excurrent into long slender mucros, lower 0,5-1, 2 mm 
(av. 0,87 mm) upper 0,27-1,04 mm (av. 0,67 mm). Wing narrow in young 


Studies in Cyperaceae in southern Africa: 10 


281 






Morphology and distribution of K. odorata. A. Spikelet with lower empty glumes re- 
moved. B. Large young golden fruitlet before maturing to black. C. Cylindrical inflo- 
rescence. D. Basal portion of leafy shoots with swollen stem-bases in cluster. 
E. Cross-sectional outline of culm. F. Distribution in southern Africa. Figure scales 
in millimetres: map scales in kilometres. 

specimens, up to 0,85 mm wide in older, ciliate, golden-green. Stamens 3. 
Style shorter than style branches. Nut 1 or 2, obovoid, yellow, turning red- 
brown when mature. 

S.W.A. — 2116 (Okahandja): Grootfontein District, Okahandja (-DD), Bradfield 
137 (PRE). 

TRANSVAAL— 2529 (Witbank): Middelburg (-CD), Gower s.n. sub PRE 29308 
(PRE); van der Merwe M.B. 07 (PRE). 

— 2530 (Lydenburg): Belfast (-CA), Leendertz 9211, 9566 (PRE). 


282 


Journal of South African Botany 


— 2626 (Klerksdorp): Hakboslaagte (-AC), Kinges 1521 (PRE). 

— 2628 (Johannesburg): Frankenwald (-AA), van Rensberg s.n. (-AD), Leendertz 
8011 (PRE). 

—2629 (Bethal): Ermelo (-DB), Collins 12223 (PRE). 

ORANGE FREE STATE — 2926 (Bloemfontein): near racecourse, Bloemfon- 
tein (-AA), Potts 3002 (NU). 

NATAL — 2632 (Bela Vista): 3,2 km from Kosi Nature Reserve (-DD), Edward 
2570 (PRE). 

— 2732 (Umbombo): Mbaswana (-BC), Michelmore 28 (PRE). 

— 2832 (Mtubatuba): Hlabisa district, Charters Creek (-AB), Ward 2776 (NH, 
PRE). 

— 2929 (Underberg): Mooi River, Meteor Ridge (-BB), Mogg 7066 (PRE). 

— 2930 (Pietermaritzburg): The Dargle (-AC), Getliffe 190 (NU); Fawn Leas (-BC), 
Getliffe 169 (NU). 

— 2931 (Stanger): Durban flats (-CC), Medley Wood 4014 (NH); Wentworth, Bluff, 
Getliffe 305 (NU), Getliffe 1008 (J). 

— 3029 (Kokstad): Harding District, Bedford Farm (-DB), Lennox s.n. (NU). 

CAPE— 3029 (Kokstad): Bizana (-DD), Strey 4336 (PRE). 

— 3125 (Steynsburg): Steynsburg (-BD), Acocks 13527 (PRE); Maraisburg district 
near Hofmeyer (-DB), Acocks 16333 (PRE). 

— 3227 (Stutterheim): Toise River (-DA), Hilner 529 (PRE). 

—3228 (Butterworth): Kei Mouth (-CB), Flanagan 927 (PRE). 

— 3326 (Grahamstown): Port Alfred (-DB), Potts 226 (BLFU). 

— 3421 (Riversdale): Riversdale (-AB), Muir 3586 (PRE). 

— 3423 (Knysna): Knysna (-AA), Breyer 23639 (PRE). 


K. alata, described as a new species by Nees (1834), was relegated to a 
variety of K. alba by Clarke (1897 and 1901). Clarke used colour of glumes, 
width of wing on the glume and pubescence of culms to distinguish these 
taxa but did not mention the difference in mucro length and shape to which 
Nees (1935-36) drew attention. 

Colour is reliable in the field where the golden spikelets of K. alata are 
conspicuously different from the white heads of K. alba , but in herbarium 
material is not always well preserved. Wing width varies with age, but is a 
reliable character in mature specimens, and pubescence of culms, particularly 
immediately below the inflorescence, is useful. Though some variation 
can be expected in K. alata in this character, no specimen K. alba was seen 
to be pubescent. While colour, wing width and pubescence may not be diag- 
nostic when used separately, these characters used in combination separate 
K. alata from K. alba. Mucro length is, however, a very useful diagnostic 
character. Using a sample of twenty specimens from each taxon from the en- 
tire range of distribution, it was established that the presence of two long, 
curved mucros per spikelet of K. alata was a consistent characteristic and 
contrasted with the single (lower), straight mucro of K. alba. 

K. alata is restricted to a southern African distribution while K. alba is 
also found in tropical Africa. 


Studies in Cyperaceae in southern Africa: 10 


283 





Fig. 9 

Morphology and distribution of K. alata. A. Terminal spike subtended by short 
bracts, on a ciliate culm. B. Stem-bases clothed in fibrous sheaths. C. Cross-sectional 
outline of culm. D. Fertile winged glumes enclosing single nutlet. E. Distribution of 
the species in southern Africa. Figure scales in millimetres: map scale in kilometres. 


284 


Journal of South African Botany 


8 . K. alba Nees, Linnaea 10: 140 (1836-36); Kunth, Enum. PI. 2: 136 
(1837), Ridl. Trans. Linn. Soc. Ser. 2, 2: 147 (1884); Clarke in Durand and 
Schinz, Conspect. FI. Afr. 5: 526 (1895), in Thiselton-Dyer, Flora Cap. 7: 
151 (1897), in Thiselton-Dyer, Flora Trop. Afr. 8 : 271 (1901), and Kew Bull. 
Addit. Ser. 8 : 93 (1908), Burtt Davy and Pott-Leendertz, Ann. Tvl. Mus: 
132 (1912); Phillips, Ann. of S.A. Mus. 16: 318 (1917); Dyer, Rec. of Alb. 
Mus. 3: 476 (1927); Schonland, S. Afr. Cyp. 19 (1922); Brain, Proc. Rhod. 
Sc. Assoc. 33: 78 (1934). Type: South Africa "in utroque latere fluminis ‘der 
zwarte Keyrivier’ dicti alt III” Ecklon (S?, holo.). 

Kyllinga cristata Kunth, Enum. PL: 136 (1837). 

Kyllinga nigritana C. B. Clarke, in Thiselton-Dyer, FI. Trop. Afr. 8 : 272 
(1902). 

K. alba Nees var. nigritana (C.B. Cl.) Podlech, Mitt. Staatssamml. 
Munch. 3: (1960). 

Cyperus cristatus (Kunth) Mattf. & Kukenth. in Engler, Pflanzenr. 4, 20: 
609 (1936). 

Cyperus cristatus (Kunth) Mattf. & Kukenth. var. nigritanus (C.B. Cl.) 
Kukenth. in Engler, Pflanzenr. 4, 20: 610 (1936). 

Kyllinga alata (Nees) Hutchinson in Hutchinson and Dalziel, Flora W. 
Tropi Afr. 1, 2: 487 (1936). 

Kyllinga cristatus var. exalatus Merxm., Proc. and Trans, of Rhod. Sc. 
Assoc. 43: 75 (1951). 

Cyperus cristatus var. exalatus Merxm. Trans. Rhod. Sc. Assoc. 43: 80 
(1951). 

K. alba var. ex alata (Merxm.) Podlech, Mitt. Staatssamml. Munch. 3: 
525 (1960). 

Plants tufted or forming a short row of sheathed stem-bases; sheaths 
brown, tough, frequently fibrous. Culms up to 520 mm tall, triangular, glab- 
rous or minutely scabrid below inflorescence. Leaves numerous, flaccid, 
generally more than half length of culm; sheaths less than 3 length of leaf; 
blade up to 5 mm wide, margin and keel scabrid especially in young leaves. 
Inflorescence of 1 (rarely 3) spikes, dense, globose to broadly elliptic, up to 
15 mm in diam., snow-white or tinged with green or discoloured by rusty 
streaks. Bracts 2-5, usually 3, patent, later pendent, up to 121 mm long. 
Spikelets numerous, up to 6,37 mm long, ovate-lanceolate in outline, fertile 
glumes 2-4, rarely more than 2 fully developed, subtending 2 florets, both 
bisexual or upper male, rarely 3 (2 bisexual, uppermost male); lower glume 
strongly 7-11-nerved, keel excurrent into straight mucro up to 0,86 mm 
long; upper 5-nerved, keel scarcely or not at all excurrent into mucro. Wing 
to keel variable, absent, narrow (approximately 0,04 mm) or wide (up to 
1 mm), glabrous or ciliate, white. Anthers 3. Style exceeded in length by 


Studies in Cyperaceae in southern Africa: 10 


285 


style branches. Nut 1, rarely 2, ellipsoid, shortly spiculate, half length of 
glume, golden, later red-brown to black. 


S.W.A.— 1713 (Swartbooisdrif): Okavango Territory, S of Omuramba (-AD) de 
Winter 3884 ( PRE). V 

—1719 (Runtu): Vlei near Runtu (-DD), Merxmtiller & Giess2135 (M)_ 

— 1729 (Sambio): East of Masare Camp (-CC), de Winter & Wiss 4100 (M, PRE). 
—1917 (Tsumeb): 50 miles W of Grootfontein (-BC), Schoenfelder 35 (PRE). 

— 1918 (Grootfontein): Grootfontein (-CA), Story 6131 (PRE) 

—2115 (Karibib): on way to Red Koppies at Karibib (-DD), Kinges 3193 (M, PRE). 
—2217 (Windhoek): Farm Otjisewa (-CA?) Wiss 755 (M); Farm Onduno (-CA?) 
Kinges 4959 (M). 


TRANSVAAL — 2231 (Pafurie): Kruger National Park S.E. of Punda Milia 
(-CA), Schieben 9525 (M). 

— 2329 (Pietersburg): Pietersburg (-CD), Codd and Dyer 9157 (PRE); Haennertz- 
burg (-DD), Pott 4751 (PRE). 

— 2330 (Tzaneen): Woodbush (-CC), Mogg 14716 (PRE). 

— 2427 (Thabazimbi): Rustenburg district, Vlakfontein (-AB), Liebenberg 158 
(PRE). 


— 2428 (Nylstroom): Nylstroom (-CB), van Dam s.n. (PRE); Naboomspruit (-DA), 
Galpin M375 (PRE). 

— 2429 (Zebediela): Schoonoord, Lydenburg district, (-DD) Mogg and Barnard 794 
(PRE). 

— 2430 (Pilgrims Rest): Pilgrims Rest (-DD), de Winter and Codd 539 (PRE). 

— 2431 (Acornhoek): Skukuza (-DC), de Winter and Codd 538 (PRE). 

— 2527 (Rustenburg): Rustenburg (-CA), Lanham 25 (PRE). 

— 2528 (Pretoria): Onderstepoort (-CA), Teiler 14805 (PRE); Pretoria, Leendertz 
7211 (PRE); do. Stent 573 (PRE); do. Bryant E2 (PRE); do. Sloffs Cutting, Smith 
177, 822 and 861 (PRE); Bronkhorstspruit (-DC), Pepton 4703 (PRE). 

— 2531 (Komatipoort): Pretoriuskop, Kruger National Park (-AB), van der Schijff 
1071 (PRE). 

— 2626 (Klerksdorp): Lichtenburg district, Hakbaslaagte (-AC), Kinges 1520 (PRE, 

M). 


— 2627 (Potchefstroom): Welverdiend (-AD), Louw 98 (PRE); Potchefstroom 
(-CA), Louw 1167 (PRE). 

— 2628 (Johannesburg): Turffontein (-AA), Bryant D50 (PRE); Heidelberg (-AD), 
Leendertz 8012 (PRE). 

— 2725 (Bloemhof): Christiana (-CC), Theron 432 (PRE); Bloemhof (-DA), Burtt 


Davy 11242, 12838, 14438 (PRE), van der Merwe 58 (PRE). 

ORANGE FREE STATE— 2727 (Kroonstad): Kroonstad (-CA), Pont 495 

(BLFU, PRE). 

— 2926 (Bloemfontein): Bloemfontein (-AA), Potts 407 (NU), Potts 3031 (NU, 
PRE), Potts 1326 (NU), Potts 3006, 3009, 3022 (BLFU); Channon (-AB), Hatchard 

7328 ( BLFU). 

— 3025 (Colesburg): Trompsburg (-BB), Potts s.n. (BLFU). 

— 3027 (Lady Grey): Pan near Quaggafontein (-AC), Potts 3006 (BLFU). 

SWAZILAND— 2531 (Komatipoort): Hlatikulu district, Havelock Concession 
(-CC), Saltmarsh 1021 (PRE). 

—2631 (Mbabane): Forbes Reef Swamp (-AA), Compton 30564 (PRE, HSS); 
Komati Pass, Mbabane (-AC), Compton 31048 (HSS); Hlatikulu, (-CD), Stewart s.n. 


(PRE). 


286 


Journal of South African Botany 


NATAL— 2632 (Bela Vista): Ndumu Hill (-CD), Ross 675 (NU); Ndumu Game 
Reserve, Ross 685, 693 (NU). 

— 2732 (Ubombo): Summit of Lebombo Mountains (-AA, -AC), Getliffe 178 (NU); 
Ubombo Village (-CA), Ward 1656 (NPGF, NU, PRE), Getliffe 162 (NU). 

— 2929 (Underberg): Warley Common (-BB), Mogg 7248 (PRE). 

LESOTHO — 2828 (Bethlehem): Caledon River, Burke 202 (PRE), Dieterlen 334 
(PRE, NH). 

CAPE— 2624 (Vryburg): Vryburg (-DC), Sharpe 7420 (PRE); do. Mogg 8170 
(PRE); Vryburg and Tiger Kloof, Brueckner 2831 (PRE); 60 miles from Vryburg 
(Farm Palmyra), Rodin 3598 (PRE). 

— 2723 (Kuruman): Kuruman, Esperanza (-AD), Esterhuysen 2143 (PRE). 

— 2823 (Griekwastad): Postmasburg (-AC), Leistner 1662 (PRE); Daniel’s Kuil 
(BA), Esterhuysen s.n. (PRE). 

— 2824 (Kimberley): Kimberley (-DB), Moran s.n. (PRE). 

K. alba is a widespread tropical and temperate species with which have 
been associated several allies. In the course of this study, each of these 
allied taxa was compared with K. alba. 

Closely allied to K. alba , according to many authors (Clarke, 1901; 
Brain, 1934; Kiikenthal, 1936 and Robinson, 1959), is K. controversa Steud. 
The original descriptions of these two species differ only in the statement 
that K. controversa has two stamens but Clarke (1901) remarked on the pale 
brown inflorescence, ciliate keelar wing and short spikelets. Studies of wing 
width have revealed considerable variation within K. alba and considerable 
overlap in range of this and allied taxa (Figure 10). This does not support 
the separation of K. alba and K. controversa on the basis of wing width. All 
specimens examined, whether tropical or southern African, had three sta- 
mens. Nor does the spikelet length, with ranges of 3,08 (av. 3,72) - 5,40 mm 
for K. alba , and 3,67 (av. 4,3) - 6,25 mm for K. controversa separate these 
two. K. controversa has therefore been placed in synonymy with 
K. alba. 

K. nigritana C. B. Clarke was distinguished by its pale cinnamon- 
coloured, broadly ovate winged spikelets. Podlech (1960) regarded it as a 
variety of K. alba. Comparison of specimens seen by Podlech with authentic 
K. alba specimens, suggests that the colour difference was the result of rusty 
streaks or dots similar to those found in gatherings of K. alba from dry local- 
ities ( Moran 1918 , Sharpe 7420, Rodin 3598). In every other respect, these 
two taxa appear identical. Similarly, K. alba var. exalata is not upheld in 
view of the variation in wing width over the range of distribution and with 
age. 

Robinson (1959) suggested that K. cartilaginea Schum. was closely allied 
to K. alba. Kiikenthal (1936) maintains it as a distinct species. Further study 
of this tropical species is necessary, but two characteristics suggest it is a sep- 
arate species: the presence of an extremely long mucro on the lower fertile 
glume giving the heads “a prickly look” (Robinson, 1959) and the very dis- 


Studies in Cyperaceae in southern Africa: 10 


287 


t 


K. alba var. exalata (Merx.) Podlech 
K. nigritana C.B.CI. 


K. controversa Steud 


K. alata Nees 


K. alba Nees 


0,3 0,5 0,7 0,8 1 


Wing width in millimetres 


Fig. 10 

Mean wing width (M), standard deviation from the mean (heavy line) and range of 
wing width (narrow line) from representative samples of K. alba and associated taxa. 


tinct rhizome morphology as evident in Napper 1371 (Herb. EARO) in 
which aerial shoots are 30-40 mm apart on an elongated rhizome. 

K. alba is therefore recognized as a complex, variable species with trop- 
ical affinities. It is a distinctive tufted species with globose white spikes 
sometimes tinged with cinnamon in drier areas. It is unlikely to be confused 
with any other South African species. 

As described above, it is distinguished from K. cartilaginea by rhizome 
characters. K. platyphylla Schum., collected rarely in Zimbabwe but fre- 
quently in Zambia and Tanzania, is similar but has broad flattened hemi- 
spherical heads composed of three confluent spikes clasped by 3-4 bracts. 

9. K. brevifolia Rottb., Descr. et Icon. 13, t.4, fig. 3 (1773); Nees in Wight, 
Contrib. Bot. India 91 (1834); Ktinth, Enum. PI. : 130 (1837); Steud., Syn- 
ops Cyper. 67 (1855); Boeck. in Linnaea 35: 424 (1868); p.p.; Clarke in 
Hooker, FI. Brit. India 6: 588 (1893), in Durand and Schinz, Conspect. FI. 
Afr. 5: 527 (1895), in Thiselton-Dyer FI. Trop. Afr. 8: 273 (1901) and in 
Illustr. Cyp. t. I Fig. 1-4 (1909); Chermezon, Catal. PI. Madag. 7 (1931), 
Barros, in Anal. Mus. argent. Cienc. Natur.: 136f .1. : 11 (1935) Hooper, in 
Hepper, Flora W. Trop. Afr. ed 2,3, 2:307 (1972). Type: India, Koenig (BM, 
holo.!; STB, iso., photostat seen). 


288 


Journal of South African Botany 



wings on the keels. B. Cross-sectional outline of stem. C. Inflorescence. D. Group of 
stem-bases. E. Distribution of the species in southern Africa. Figure scales in milli- 
metres: map scales in kilometres. 


Studies in Cyperaceae in southern Africa: 10 289 

Cyperus brevifolius (Rottb.) Hassk., Kiikenth., in Engler Pflanzenr 
20:600 (1936). 

K. cruciata Nees in Linnaea 9: 286 (1834), name only. 

Rhizome slender (1-2| mm diam.), elongate, white or greenish-white, 
covered with membranous, striate scales. Culms 5-30 mm apart, slender, 
7—32 mm tall, triangular, glabrous, bases unthickened. Leaves few, equal- 
ling or shorter than culms; sheaths membranous, lower 1-2 leafless or 
briefly laminate; blades 1-3 mm wide, margins and keel scabrid, shortly acu- 
minate. Bracts 3-4, long, erect, later spreading. Inflorescence of 1 spike 
(rarely 2 or 3), ovoid to ellipsoid, greenish-white, dense. Spikelets numer- 
ous, tightly packed, elliptic in outline, 2-3,5 mm long; floret 1, bisexual, 
rarely 2, lower bisexual, upper male. Fertile glumes 2, membranous, trans- 
parent white, lower 7-nerved. upper 3-5-nerved, keel green, scabrid, excur- 
rent into acute recurved mucro (0,2-0,45 mm). Stamens 1 per floret. Style 
short, far exceeded in length by style branches. Nut \ length of glume; 
broadly obovate, yellow. 

NATAL — 2930 (Pietermaritzburg): Vlei in Pietermaritzburg (-CB), Thurgood 
s.n. (NU); Chase Valley, Getliffe 23 (NU); Wyllie Park, Ross 357 , 358 (NU); Uni- 
versity of Natal, Getliffe 191, 228 (NU); Botanical Gardens, Ross 1266 (NU); Golf 
Road, Pietermaritzburg. Ross 1267 (NU); King Edward Avenue, Getliffe 998, 999 
(J): Thomas Moore School. Pinetown (-DD), Getliffe 1003 (J); Edgewood Training 
College. Pinetown, Getliffe 1007 (J). 

— 2931 (Stanger): Jameson Park (-CC), Hilliard s.n. (NU); Windsor Park Golf 
Course, Getliffe 976 (J). 

Although it appears, as K. erecta , in Henderson and Anderson (1966) as 
a weed in South Africa, there has been no other record of this species from 
southern Africa until the present study. It has, however, been well recorded 
for west tropical Africa (Clarke, 1901; Hutchinson, 1936; Hooper, 1972b). 
The possibility of this frequent weed of lawns being an introduction cannot 
be dismissed. 

There appears to be some confusion in the older descriptions over the 
stamen number. Linnaeus in his Species Plantarum placed it among plants 
with three stamens, Kunth (1837) agreed but Barros (1935) working on 
South American material recorded one stamen only per floret. Kiikenthal 
(1936) gave a range of 1-3. Either this is an extremely variable taxon, or 
several distinct taxa have been confused in the description. South African 
specimens have but a single stamen per floret. 

Apart from this distinctive feature, it is recognized by its slender white 
rhizomes (without the striking red-brown scales of K. intricata ) and its 
greenish-white ovoid to ellipsoid spikes. It cannot be confused with K. erecta 
which has a fleshy rhizome and contiguous, bulbous-based stems. 


290 


Journal of South African Botany 




Morphology and distribution of K. brevifolia. A. Spikelet with lower sterile bract (i); 
and prophyll or empty glume (ii) and two glumes enclosing 1 floret with a single sta- 
men. B. Young nutlet. C. Inflorescence with one erect bract and two smaller spread- 
ing bracts. D. Cross-sectional outline of stem. E. Portion of slender elongated rhi- 
zome. F. Distribution of the species in Natal. Figure scales in millimetres: map scale 

in kilometres. 


10 . K. elatior Kunth, Enum. PI. 2: 135 ( 1837 ); Steud., Synops. Cyper. 69 
( 1855 ); Boeck in Linnaea 35: 422 ( 1868 ); C. B. Clarke, in Durand and 
Schinz, Conspect. FI. Afr. 5: 528 ( 1895 ), in Thiselton-Dyer, FI. Cap. 7: 153 
( 1897 ), in Thiselton-Dyer, Flora Trop. Afr. 8: 275 ( 1901 ) and in Kew Bull. 
Add. Ser. 8: 93 ( 1908 ); Medley Wood in Trans. S.A. Phil. Soc. 2: 250 
( 1908 ); Chermexon, Catal. PI. Madag. 7 ( 1931 ); Hooper in Hepper, Flora 
W. Trop. Afr. ed 2 , 3, 2:307 ( 1972 ). Type: South Africa, Pondoland. Drege 
4384 (B, holo., possibly destroyed; P, iso.!). 


Studies in Cyperaceae in southern Africa: 10 


291 


Cyperus aromaticus (Ridl.) Mattf. et Kiikenth. var. elatior (Kunth) 
Kiikenth. in Engler, Pflanzenr. 4, 20:582 (1936). 

Roots thick. Rhizome elongated, horizontal, slender (2-3 mm diam.), 
pliable, covered with acuminate, striate scales with transparent membranous 
margins. Stem-bases scarcely thickened, covered with scales and pinkish- 
brown sheaths. Culms 150-600 mm tall, triangular, distant, 3-22 mm apart, 
average 17 mm apart, glabrous. Leaves few, shorter than culms, lower 
leaves reduced to sheaths, upper 2-4 sheaths laminate; sheaths membra- 
nous, pinkish-brown, frequently transversely wrinkled below mouth; blades 
2-6 mm wide, scabrid on margins and keels and on lower leaf surface, par- 
ticularly near tip. Bracts 4-6, spaced at short intervals below inflorescence, 
up to 290 mm long, spreading at right angles to culms. Inflorescence of 3-1 
green spikes, dense; central spike cylindric, 7-20 mm long, 5-13 mm wide; 
lateral smaller. Spikelets numerous 2, 7-4, 5 mm long, ovate-lanceolate in 
outline; fertile glumes 3; lower glume 7-nerved, subtending bisexual floret, 
middle 5-nerved, subtending male (rarely bisexual) floret; keels green, wing- 
less, glabrous or with 1-4 hairs, excurrent into slender, acuminate, recurved 
mucro, 0,28-0,59 mm long. Stamens 3, anthers linear, connective slightly 
produced and red. Style greatly exceeded in length by branches. Nut (rarely 
2), obovoid, approximately 1 mm long, red-brown, later black. 


NATAL — 2831 (Nkandhla): Ngoye Forest Reserve, streambank (-DC), Getliffe 
215 (NU); Qwesqweqneqnolo, Ngoye Forest Reserve, Getliffe 217 (NU); Forest 
Guard Hut, Nogoye Forest Reserve, Getliffe 220 (NU); Mtunzini, Port Durnford 
Forester's Office (-DD), Getliffe 214 (NU). 

— 2832 (Mtubatuba): Richards Bay (-CC), Ward 714 (NU); do. Getliffe 60 (NU). 

— 2930 (Pietermaritzburg): Lions River district, “The Start”, Karkloof (-AD), Get- 
liffe 205 (NU); Town Bush Valley, Pietermaritzburg (-CB), Ward 118 (NU); do. Get- 
liffe 196 (NU); Table Mountain, Killick 375 (NU, PRE); Chase Valley, Pietermaritz- 
burg, Jolliffe 34 (NU); do. Wilson 11 (NU); do. Getliffe 25 (NU); do. Ross 728 
(NU); Enon Estate, Richmond (-CD), Getliffe 165 (NU); Nagle Dam (-DA), Wells 
1021 (NU); Ichanga (-DC), Armour s.n. (NU); Isipingo Flats (-DD), Ward 4961 
(NU); Wentworth, relict of Bayhead Swamps (-DD), Ward 4971 (NU). 

—2931 (Stanger): Farm Waterbosch, Stanger (-AD), Bodhoo s.n. (NU); Stanger 
Secondary School, Getliffe 226 (NU); Inanda district, Mount Edgecombe (-CA), Mc- 


Martin 9 (NU); do. Getliffe 201 (NU); 

— 3030 (Port Shepstone): Isipingo Beach (-BB), Ward 4065 (NU); do. Ward 4968 
(NU); Sezeta River (-BC), Getliffe 31 (NU); Glenmore, Port Shepstone district 
(-CD), Anderson 55 (PRE). 

CAPE— 3129 (Port St. Johns): Port St. Johns district (-DA), Schonland 4162 
(PRE) 

—3227 (Stutterheim); NW of Zele P.O., King Williamstown (-CD), Acocks 20279 


(PRE, SRGH). 


This elegant species with long horizontal bracts and elongate slender rhi 
zomes bearing culms at wide intervals occurs in damp shady situations where 


292 


Journal of South African Botany 


it may form dense stands. The inflorescence is a cylindrical spike borne on 
acutely three-angled stems. Not only is the rhizome fragrant but the aerial 
parts are delicately lemon-scented when bruised. A mat, possibly for carry- 
ing maize meal and known as “isitebe”, is made from these culms (Medley 
Wood, 1908). 

K. elatior is unlikely to be confused with any other South African species 
but the description in Brain (1934) is misleading being based on an incor- 
rectly named gathering of K. melanosperma. The long bracts and elongate 
rhizome are reliable features distinguishing these species; similarly the con- 
gested rhizome of K. polyphylla and its composite inflorescence of globose 
spikes distinguishes this from K. elatior. 

11. K. intricata Chermezon, in Bull. Mus. Hist. Nat (Paris) 25: 211 (1919); 
Catal. PI. Madag. 8 (1931); Podlech in Prodr. FI. Siidwestafr. 165:31 (1967). 
Type: Madagascar, Hildebrandt 3788 (K, holo.). 

K. erecta var. intricata C. B. Clarke in Durand and Schinz, Conspect. FI. 
Afr. 5: 529 (1895). 

Cyperus erectus (Schum.) Mattf. and Kukenth. var. intricatus (C. B. 
Clarke) Kukenth. in Engler, Pflanzenr. 4 (20): 590 (1936). 

Rhizome elongate, slender, 1-2 mm diam., horizontal, covered with loose, 
red-brown scales. Culms 5— (15)— 30 mm apart, bases not thickened. Leaves 
2-9 per culm, shorter than culms. Bracts 3, lowest erect, later reflexed. 
Inflorescence 1 globose spike 4-8 mm wide, compact golden-buff. Spike- 
lets many, ovate — lanceolate in outline, 2-4 (av. 2,97) mm long. Fertile 
glumes 2 or 3, briefly mucronate, lower mucro av. 0,18 mm, upper 0,13 mm. 
Style shorter than branches. Nut 1, less than half length of glume, ellipsoid 
to obovoid, yellow to black. 

S.W.A. — 1719 (Runtu): Okavango Reserve (-DD), de Winter 3744 (M, PRE). 

— 1821 (Andara): Caprivi Zipfel, Andara (-AB), Merxmiiller & Giess 1977 (M); E 
Caprivi (-AB?), Killick & Leistner 3034 (PRE). 

TRANSVAAL — 2528 (Pretoria): 15 km from Hammanskraal (-AD), Kies 374 
(PRE); do. Codd 3512 (PRE). 

NATAL — 2931 (Stanger): Verulam (-CA), Medley Wood 1076 (NH). 

K. intricata is a tropical species with limited distribution in South Africa. 
It is allied to K. erecta but its distinctive rhizome with elongated internodes 
and loose overlapping scales distinguish it from that species. The mucro 
length in K. intricata is distinctly shorter than that of K. erecta , spikelets are 
also smaller on an average but the ranges overlap making this less valuable 
than rhizome and mucro features. 

12. K. erecta Schumach. in Schum. et Thon., Besk. Guin. PI. 63 (1827); C. 
B. Clarke in Durand and Schinz, Conspect. FI. Afr. 5: 528 and in Thiselton- 


293 


Studies in Cyperaceae in southern Africa: 10 



Fig. 13 

Morphology and distribution of K. elatior. A. Inflorescence subtended by long 
spreading bracts. B. Robust elongated rhizome clothed in striate scales. C. Cross- 
sectional outline of culm. D. Spikelet with lower emply glumes (bracts and prophyll) 
and two fertile glumes. E. Floret from within fertile glumes. F. Distribution in south- 
ern Africa. Figure scales in millimetres: map scale in kilometres. 


Dyer, Flora Trop. Afr. 8:274 (1901); Rendle, in Catal. Afr. PI. Welw. 2: 
105 (1899); Medley Wood, in Trans. S.A. Phil. Soc. 18, 2: 250 (1908); 
Schonland, in Mem. bot. Surv. S. Afr. 3: 20 til (1922); Chermezon. Catal. 
PI. Madag. 7 (1931), Hutchinson in Hutchinson and Dalziel, Flora W. Trop. 
Afr. 2, 2: 487 (1936): Nelmes and Baldwin, in Amer. Journ. Bot. 39: 389 
(1952); Hooper in Hepper, Flora W. Trop. Afr. ed 2, 3, 2: 307 (1972). 
Type: Guinea, Isert and Thonning (C). 


294 


Journal of South African Botany 



Morphology and distribution of K. intricata. A. Spikelet including empty lower 
glumes (B). C. Inflorescence with one erect bract. D. Portion of rhizome clothed in 
loose brown scales. E. Distribution in southern Africa. Figure scales in millimetres: 
map scale in kilometres. 

K. aurata Nees in Linnaea 10: 139 (1935-36), Kunth, Enum. PI. 2: 137 
(1837), Steud., Synops Cyp. 69 (1855), Boeck. in Linnaea 35: 422 (1868), 
Ridl. in Trans. Linn. Soc. 2 ser. Bot. 2: 146 (1884). 

K. colorata Hutchinson, in Hutchinson and Dalziel, Flora W. Trop. Afr. 
2, 2: 487 (1936) (non (L.) Druce). 

Cyperus erectus (Schum.) Mattf. and Kiikenth. in Engler, Pflanzenr. 4, 
20: 588 (1936). 

C. erectus var. aurata (Nees) Kiikenth. in Engler, l.c. 589 (1936). 

Rhizome moderately long, of bulbous stem-bases, internodes short, 
clothed in acuminate, red-brown, seldom overlapping, scales. Culms pro- 
duced in uniseriate row, slender, triangular to rounded, glabrous, bases bul- 
bous, thickened. Leaves few, shorter than culms; sheaths short, lowest 2-3 
leafless, red-brown, blade flat, 2-3 mm wide, flaccid, acuminate. Bracts 3, 
lowest erect, later reflexed. Inflorescence of 1 spike, globose to ovoid, 4-8 
mm wide, compact, greenish-yellow to red-golden. Spikelets many, ovate- 


Studies in Cyperaceae in southern Africa: 10 


295 


lanceolate in outline, 2-3 mm long, fertile glumes 2 or 3 golden, frequently 
rust-streaked, lowest prominently 7-9-nerved, subtending bisexual flower, 
middle 3-5-nerved, subtending male flower, sometimes enclosing reduced 
hyaline uppermost empty glumes, keels green when fresh, glabrous or with 1- 
2 scabrid hairs, excurrent into acute recurved mucro 0,2-0, 5 mm long (av- 
erage length of lower mucro 0,35 mm). Style exceeded in length by style 
branches. Nut 1, less than \ length of glume, ellipsoid to obovoid, yellow to 
black. 

TRANSVAAL — 2530 (Lydenburg): Machadodorp (-CB), de Winter and Codd 
227 (PRE). 

— 2531 (Komatipoort): Pretoriuskop (-AB), van der Schijf 1999 (PRE). 

— 2626 (Klerksdorp): Hakboslaagte (-AC), Kinges 1648 (PRE). 

— 2627 (Potchefstroom): Potchefstroom district, Laerberg (-CA), Theron 686 (PRE); 
Potchefstroom, Louw 1216 (PRE). 

— 2628 (Johannesburg): Frankenwald (-AA), Gilliland 25059 (PRE). 

— 2629 (Bethal): Ermelo (-DB), Leendertz 10539 (PRE); Noodgedacht, Potter 1781 
(PRE). 

ORANGE FREE STATE— 2828 (Bethlehem): Bethlehem (-AB), Phillips 3102 
(PRE). 

NATAL — 2632 (Bela Vista): Ndumu Game Reserve (-CD); Ross 659, 687, 694 
(NU). 

— 2732 (Ubombo): False Bay Park (-CD), Ward 4144 (PRE). 

— 2832 (Mtubatuba): Hluhluwe Game Reserve (-AA), Ward 1793 (NH, NPGF, NU, 
PRE); St. Lucia Park (-AD), Dutton 62 (NU). 

— 2930 (Pietermaritzburg): New Hanover, Fawnleas (-BC), Getliffe 170 (NU); Town 
Bush Valley (-CB), Ward 110 (NU); Oribi, Evans 17 (NU); Thomas Moore School 
(-DD), Getliffe 176 (NU). 

— 2931 (Stanger): Mount Edgecombe (-CA) Most 77 (NU). 

CAPE— 2624 (Vryburg): Moita, Mafikeng District (-BB), Brueckner 554 (PRE). 
— 3326 (Grahamstown): Grahamstown (-BC), Daly & Sole 202 (PRE); Trappes Val- 
ley, Daly 12934 (PRE); do. Lubke 88 (NU); Alexandria District near Whitney 
(-DA), Archibald 5068 (PRE). 

— 3327 (Peddie): between Core Rock and Gooda River (-BB), Galpin 7355 (PRE). 

— 3420 (Bredasdorp): Heidelberg (-BB), Esterhuysen 14455 (PRE). 

Kyllinga erecta Schum. and K. aurata Nees were regarded as synonyms 
by Clarke (1897) and Schonland (1922), but Kiikenthal recognized them as 
two varieties of Cyperus erectus. He distinguished the variety aurata by its 
golden to rust forms, shorter culms, narrower leaves and longer mucronate 
tips to the keel of the glumes. 

Height of the plant cannot be regarded as diagnostic as it is so dependent 
upon environmental conditions and indeed Dieterlen 211 (which Kiikenthal 
cites as the variety aurata ) exceeds his 200 mm limit for C. erectus var. aura- 
tus. Leaf-width has been shown to be unreliable and no significant difference 
in mucro length could be detected. 

In the more southerly collections the colour of the head appears to be 
prominently golden but in the absence of any clear-cut distinction it is pro 
posed that K. erecta and K. aurata are conspecific. 


296 


Journal of South African Botany 



Fig. 15 

Morphology and distribution of K. erecta. A. Whole spikelet. B. Inflorescence with 
erect bract. C. Rhizome comprising short swollen internodes with bulbous-based 
shoots not contiguous. D. Cross-sectional outline of culm. E. Distribution of species 
in southern Africa. Figure scales in millimetres: map scale in kilometres. 

13. K. melanosperma Nees in Linnaea 9: 286 (1834) name only, in Wight, 
Contributions to the Bot. of India, 91 (1834); Kunth, Enum. PI. 2: 131 
(1837); Steud. Synops. Cyper. 67 (1855); Boeck. in Linnaea 35: 419 (1868); 
Oliver in Trans. Linn. Soc. 29: 166 (1875); C. B. Clarke, in Hooker FI. Brit. 
India 6: 588 (1893), in Durand and Schinz, in Thiselton-Dyer, Flora Cap. 7: 
153 (1897), in Thiselton-Dyer, Flora Trop. Afr. 8: 277 (1901) and in Bull. 
Misc. Inf. Add. Ser. 8: 93 (1908); Medley Wood, in Trans. S.A. Phil. Soc. 
17, 2: 250 (1908); Burtt Davy & Pott-Leendertz, in Ann. Trans. Mus. 132 
(1912); Schonland in Mem. bot. Surv. S. Afr. 3: 20 (1922); Dyer, in Rec. 
Albany Mus. 470 (1927); Fourcade in Mem. bot. Surv. S. Africa 20: 85 
(1941); Hooper in Hepper, Flora W. Trop. Afr. ed 2, 3, 2: 307 (1972). 
Type: India, Wight 1851 (K, holo.). 

Cyperus melanospermus (Nees) Kiikenth. in Engler, Pflanzenr. 4, 20: 
583 (1936). 

Roots thick. Rhizome horizontal, stout (4-8 mm diam.), congested (in- 
ternodes 3-5 mm) covered with short red-brown, acuminate, imbricate 
scales with hyaline margins. Stem-bases bulbous-thickened, covered with 
cuspidate scales. Culms 150-1 000 mm tall, firm, triangular, glabrous, con- 


Studies in Cyperaceae in southern Africa: 10 


297 


gested, in uniseriate rows or massed in groups. Leaves few, lower leaves re- 
duced to red-brown membranous sheaths, upper 2-4 sheaths shortly lami- 
nate; sheaths long, mouth of sheath often membranous and transversely 
wrinkled; blade 3-4 mm wide, flat, scabrid on margins and keels, stiff. 
Bracts 3 or 3 plus 1 very small bract, relatively short (up to 120 mm), 
upright, later reflexed, stiff. Inflorescence a single spike, ovoid-cylindric to 
subglobose, 5-14 mm long by 6-10 mm wide, dense golden-green (rust- 
coloured under unfavourable conditions). Spikelets numerous, ovate-lanceo- 
late in outline, 3-5,5 mm long (average length 3,97 mm), fertile glumes 2, 
unequal, plus third upper reduced hyaline empty glume, lower glume 7- 
nerved, subtending bisexual floret, middle 5-nerved, subtending male, rarely 
bisexual floret; keels green, wingless, glabrous or sparsely scabrid, excurrent 
into slender, acuminate, slightly recurved mucro (0,25-0,70 mm long). Sta- 
mens 3, anthers linear, connective slightly produced and red. Style far ex- 
ceeded in length by style branches. Nut 1 (rarely 2), small (approximately 1 
mm long), obovoid, golden turning bright red-orange later black, slightly 
apiculate. 


BOTSWANA— 2525 (Mafikeng): Kanye district, source of Ramongola river 
(-AB), Hilliary and Robertson 576 (PRE). 

TRANSVAAL — 2288 (Maasstroom): 2 miles E of Donderhoek (-DD), Killick 
1533 (NU). 

— 2329 (Pietersburg): Haenertsburg (-DD), Pott 13641 (PRE). 

— 2330 (Tzaneen): Westfalia estate, Duiwelskloof (-CA), Scheepers 167 (PRE); 
Magoesbaskloof (-CC), Winter & Codd 135 (PRE); do. Codd 3076 (PRE); Tzaneen, 
Phillips 3285 (PRE). 

— 2428 (Nylstroom): Warmbaths (-CD), Leendertz 6199 (PRE). 

— 2527 (Rustenburg): Rustenburg (-CA), Leendertz 3428 (PRE). 

—2528 (Pretoria): Aapies River (-CA), Leendertz 6216 (PRE); Pretoria, Repton 734 


(PRE). 

—2529 (Witbank): near Groblersdal (-AB), Acocks 20901 (PRE). 

— 2530 (Lydenburg): Waterfall (-CB), Rogers 14398 (PRE). 

—2531 (Komatipoort): Nelspruit (-AD), Breyer 17978 (PRE); Pretoriuskop, van der 
Schijffl981 (PRE). 

— 2627 (Potchefstroom): Krugersdorp, Farm Gladysvale (-BB), Rodin 3921 (PRE). 
— 2628 (Johannesburg): Rietvallei Farm, Johannesburg (-AA), Mogg 14769 (PRE), 
Frankenwald, Rose Innes 7 (PRE). 

— 2630 (Carolina): Marimstad (-CA), Pott 5226 (PRE). 

SWAZILAND— 2631 (Mbabane): Black Mbuluzi Valley Swamp (-AA?), Comp- 
ton 28258 (PRE); Marikiana Swamp (-CA), Compton 28460 (HSS, PRE); near Hla- 
tikulu on Goedegun Road (-CD), Ross 1477 (NU). 

— 2731 (Louwsberg): Goedgegun (-AA), Ross 1430 (NU). 

NATAL— 2730 (Vryheid): Pongola River (-AD), Getliffe 285 (NU ); 5 on i Louws- 
berg Road near Vryheid (-DD), Getliffe 283 (NU); do. Ross 1212 (NU), do. oss 

— 2830%undee): Klip River District, near Wessels Nek (- AC) , Pentz^ and. Acocks 
A10269 (PRE); near Nqutu on Qudeni Road (-BA), Edward 2190 (PRE); Mtunzmi, 
Hamewith (-CC), Mogg 6028 (PRE). 


298 


Journal of South African Botany 


— 2929 (Underberg): Estcourt district, Blaauwkrantz River (-BB), Getliffe 267 (NU); 
Bergville (-CD), Edwards 2357 (PRE). 

— 2930 (Pietermaritzburg): Lidgetton (-AC), Mogg 801 (PRE); Karkloof, Gordon- 
Gray s.n. (NU); Howick, Getliffe 223 (NU); Albert Falls (-AD), Dement 28 (NH, 
NU); Town Bush Valley, Getliffe 309 (NU); Greytown (-BA), Medley Wood 641 
(NH), Getliffe 223 (NU); Lilani (-BB), Hilliard s.n. (NU); Hermannsburg, Getliffe 
173 (NU). 

— 2931 (Stanger): flats near Durban (-CC), Medley Wood 4100 (NH). 

— 3030 (Port Shepstone): near Umtamvuna Pont (-CC), Getliffe 30 (NU). 

District unknown: Kumbula Mission Station, Gerstner 4333 (PRE). 

CAPE — 3227 (Stutterheim): Stutterheim (-CB), Acocks 9240 (PRE). 

K. melanosperma has been widely collected in southern Africa no doubt 
partly because it is conspicuous in the field forming dense stands near water. 
The species may be recognized by its stout congested rhizomes and the 
cylindrical head subtended by short bracts which reflex from erect position 
to an angle of 45 ° from the culm when older. 

Gerstner, in notes accompanying his specimen number 4333 (from Kwa- 
Zulu), mentioned that culms were used by the Zulus to make “in Xosahne” 
a sieve for native beer. Grant (in Oliver, 1875) described K. melanosperma 
as a “sweet smelling plant”, the roots of which were pounded by Wezee 
women and rubbed on their skin as a scent. Their name for this species is 
“Keetolo” (Grant, l.c.). 

Schonland (1922) pointed out that Clarke (1901) distinguished inad- 
equately between K. melanosperma and K. elatior, having neglected rhi- 
zome structure which I believe to be an extremely valuable diagnostic 
character. Very similar in gross morphology is K. polyphylla which is de- 
scribed below. 

14. K. polyphylla Willd. ex Kunth, Enum. PI. 2, 134 (1837); Boeck., in Lin- 
naea 35: 409 (1868); C. B. Clarke in Durand and Schinz, Conspect. FI. Afr. 
5: 531 (1895), and in Thiselton-Dyer, Flora Trop. Afr. 8: 276 (1902) and in 
Kew Bull. Add. Ser. 8: 93 (1908) (name only); Chermezon, Catal. PI. 
Madag. 8 (1931). Type: India, Herb. Willd 1441 (B. holo. — seen in photo- 
graph). 

Kyllinga macrantha Boeck. in Linnaea 35: 420 (1868) p.p. 

Kyllinga aromatica Ridl. in Trans. Linn. Soc. 2, ser. 2 Bot.: 146 (1884). 
Kyllinga planiceps C. B. Clarke in Durand and Schinz, Conspect. FI. 
Afr. 5: 531 (1895). 

Kyllinga elatior C. B. Clarke in Durand and Schinz, l.c.:528 pro parte 
(Angolan material), Rendle in Catal. Afr. PI. Welw. 2: 104 (1899). 

K. erecta Schum. var. polyphylla (Willd. ex Kunth) Hooper, Kew Bull. 
26: 580 (1972), Hooper in Hepper, FI. W. Trop. Afr. ed 2. 3, 2: 307 (1972). 

Cyperus aromaticus (Ridl.) Mattf. et Kiikenth. in Engler, Pflanzenr. 4, 
20: 581 (1836). 


Studies in Cyperaceae in southern Africa: 10 


299 



Morphology and distribution of K. melanosperma. A. Inflorescence subtended by 
short stiff bracts. B. Cross-sectional outline of the culm. C. Whole spikelet with rust 
streaks on flanks of fertile glumes. D. Congested contiguous stem-bases forming rhi- 
zome. E. Distribution of the species in southern Africa. Figure scales in millimetres: 
map scale in kilometres. 


300 


Journal of South African Botany 


Roots thick, fibrous. Rhizome horizontal, stout (5-8 mm diam.), con- 
gested (internodes 2-4 mm), covered with red to black acuminate, imbricate 
scales with transparent membranous margins. Stem-bases bulbous-thick- 
ened, covered with red-brown cuspidate scales with membranous margins. 
Culms 300-640 mm tall, triangular, glabrous, congested, usually in unise- 
riate row. Leaves 2-4, far apart, less than | length of culms, lower leaves re- 
duced to sheaths, upper sheaths shortly laminate; sheath long, membranous, 
red-brown, blade flaccid, 2-4 mm wide, flat, margins on keel scabrid. Bracts 
3-8, usually approximately equal in length, 130-210 mm spreading. Inflores- 
cence of 3 spikes confluent into subglobose head, greenish-white, central 
spike 5-10 mm in diam., lateral much smaller. Spikelets numerous, ovate- 
lanceolate in outline, 3-4 mm long. Fertile glumes 2, unequal, plus third 
hyaline reduced upper glume, lowest 9-nerved, subtending bisexual floret, 
middle 5-nerved, subtending male floret, upper empty; glume keels green, 
wingless, sparsely spinulose, excurrent into slender acuminate recurved 
mucro (0,20-0,57 mm long). Stamens 3, anthers linear, connective produced 
and red. Style exceeded in length by style branches. Nut 1 small (approxi- 
mately 1 mm long), obovoid, apiculate, golden to red-brown, later black. 

NATAL — 2931 (Stanger): Durban, roadside Windsor Avenue, Windsor Park 
Golf Course (-CC), Getliffe 28, 164, 282 (NU). 


In May, 1963, a population of Kyllinga plants was found growing in light 
sandy soil on the roadside in Windsor Avenue, Durban. This was a well 
established but very localised population. Plants were characterised by the 
3-6 “flaccid” bracts aggregated at the apex of each culm and subtending a 
very small inflorescence usually composed of three closely associated, green- 
ish-white, globose heads. Subsequent identification showed these plants to 
be K. polyphylla Willd. ex Kunth. Confirmation was obtained by compari- 
son of dried specimens with sheets in the British Museum. 

Mr. J. Lewis who carried out these comparisons, and to whom grateful 
thanks are due, commented as follows, “The British Museum has many 
sheets of this species bearing determinate labels K. polyphylla in Clarke’s 
hand. These match Getliffe specimens very well.” 

Included amongst the sheets with which the Getliffe specimens were 
compared was the type of K. planiceps C.B. Clarke, a later synonym of K. 
polyphylla. 

This single record of K. polyphylla is intriguing and difficult to explain, 
and despite extensive collection along the Natal coast, no other populations 
have been found. This then is the southern-most record of this tropical 
species. 


Studies in Cyperaceae in southern Africa: 10 


301 





moved (B). C. Inflorescence of major terminal head and two minor lateral heads, 
with subtending bracts well developed. D. Cross-sectional outline of culm. E. Con- 
gested stem-bases constituting the rhizome. F. Known distribution in South Africa re- 
stricted to date to a single record in Natal. Figure scales in millimetres: map scale in 

kilometres. 


302 


Journal of South African Botany 


Hooper (1972) related K. polyphylla to a variety of K. erecta which deci- 
sion is not retained here as K. erecta is very different in size and orientation 
of bracts, rhizome morphology, and inflorescence form. It is also very dis- 
tinct from K. elatior but could be confused with K. melanosperma unless at- 
tention was drawn to the long bracts and multiple head. The rhizome of K. 
polyphylla in Durban is always obliquely ascending but this may be due to 
the adjustment due to topdressing on the grassy verge. 

General 

The genus Kyllinga includes grassland forms such as K. odorata, K. alba 
and K. alata which appear to be able to tolerate full sun and drier soil than 
the majority of the genus which exhibits a preference for damp habitats. 
K. melanosperma is intolerant of shading by surrounding vegetation but 
K. elatior thrives in shady situations. In an experiment testing the range of 
tolerance of soil moisture changes, K. nemoralis proved to be extremely 
adaptable though in natural populations it was always found in damp sandy 
soils. 

The centre of distribution of the genus is undoubtedly tropical Africa. 
Migration southwards appears to have taken place along two flanks. The 
western path leads to a distribution in tropical Africa extending into South 
West Africa as is the case in K. albiceps and K. welwitschii. The eastern 
path is more extensive. Some species penetrate as far as Natal and are re- 
stricted to coastal areas ( K . polyphylla and K. elatior). Others are more ex- 
tensive and occur throughout the Transvaal, Orange Free State and Natal 
( K . melanosperma, K. odorata, K. pauciflora, K. erecta, K. intricata). Only 
K. alba has been recorded from tropical Africa and from across Botswana to 
the east coast and south through the northern Cape, Orange Free State and 
Natal. K. alata , in contrast, is limited to southern Africa and has been re- 
corded in the south western Cape. 

Acknowledgements 

The author is indebted to the University of Natal for use of the facilities 
of the Bews Botanical Laboratories. The generous help and guidance of 
Professor Kathleen Gordon-Gray, Mr. John Lewis (British Museum of Na- 
tural History), Miss Sheila Hooper (Kew) and Mr. E. A. Robinson is grate- 
fully acknowledged, as is the unfailing enthusiasm and support of her family. 
The cooperation of the directors of herbaria who provided specimens includ- 
ing type material is acknowledged with thanks: 

Botanischer Garten und Botanisches Museum, Berlin-Dahlem; 

British Museum of Natural History; 

Botanical Museum and Herbarium, Copenhagen; 

Moss Herbarium, University of Witwatersrand; 


Studies in Cyperaceae in southern Africa: 10 


303 


Herbarium, Royal Botanic Gardens, Kew; 
National Herbarium of Victoria, Melbourne; 
Natal Herbarium, Durban; 

University of Natal Herbarium; 

Fielding-Druce Herbarium, Oxford; 

Museum National d'Histoire Naturelle, Paris; 
Botanical Research Institute, Pretoria; 

Swedish Museum of Natural History, Stockholm; 
Bergius Foundation, Stockholm. 


References 

Allan, R. D., Wells, R. J., Correll, R. J. and MacLeod, J. K., 1978. The pres- 
ence of quinones in the genus Cy perns as an aid to classification. Phyto- 
chem. 17: 263-266. 

Barros, M., 1935. Ciperaceas Argentinas. An. Mus.' argent. Cienc. nat. 38: 133. 
Bentham, G. and Hooker, J. D., 1883. Genera Plantarum 3: 1045. 

Boeckeler, J. O., 1868. Die Cyperaceen des koninglichen Herbarium zu Berlin. 
Linnaea 35: 397-434. 

Brain, C. K., 1934. A key to the Sedges of Southern Rhodesia. Proc. Trans. Rhod. 
scient. Ass. 33: 78. 

Clarke, C. B., 1893. In: Hooker, J. D., Flora of British India 6: 587-589. London: 
Reeve & Co. 

, 1895. In: Durand and Schinz, Conspectus Florae Africae 5: 526-534. 

, 1897. In: Thiselton-Dyer, Flora Capensis 7: 151-155. 

, 1901. In: Thiselton-Dyer, Flora of Tropical Africa 8: 268-288. 

, 1908. New genera and species of Cyperaceae. Kew Bull. Additional Series 

8: 93 (Kew). 

, 1909. Illustration of Cyperaceae. Tab. 1 and 2. London: Williams and 

Norgate. 

Dandy, J. E., 1935. Additions and amendments to International Rules of Botanical 
Nomenclature. Kew Bull. 83. 

Engler, H. G. A. and Prantl, K. A. E., 1887. Nat. Pflanzenfam. 2, 2: 98-126. 
Leipzig: Engelman. 

Getliffe, F. M. and Baijnath, H., 1976. Studies in Cyperaceae in southern Africa: 
7. Cy perus prolifer Lam. A case of mistaken identity. Jl S. Afr. Bot. 42 (2): 
273-281. 

Henderson, M. and Anderson, V. G., 1966. Common weeds in South Africa. 
Mem. bot. Surv. S. Afr. 37. 

Hooper, S., 1972a. New taxa names and combinations in Cyperaceae for the Flora 
of West Tropical Africa”. Kew Bull. 26: 577-583. 

, 1972b. In: Hepper, N., Flora W. Trop. Afr. ed 2. 

Kern, J. H., 1974. Cyperaceae. In: van Steenis, C. G. G. J., Flora Malesiana, ser. 1, 
7: 435-753. Alphen. 

Koyama, T., 1961. Classification of the family Cyperaceae. J. Fac. Sci. Tokyo Univ. 
Ser. 3, 8: 36-148. 


304 Journal of South African Botany 

Kukenthal, G., 1936. In: Engler, H. G. A., Das Pflanzenreich 4, 20: 566-614. 
Stuttgart. 

Kunth, C. S., 1837. Enumeratio Plantarum 2: 127-138. Stuttgart and Tiibinger. 
Lanjouw, J. (ed.), 1978. I.C.B.N. Regnum veg. 97: 303. 

Linnaeus, C., 1753. Species plantarum. ed. 1. 

Napper, D. M., 1971. Cyperaceae of East Africa. Jl. E. Africa nat. Hist. Soc. 28. 
124: 1-24. 

Nees, C. G.: see Nees ab Essenbeck, C. G. 

Nees ab Essenbeck, C. G., 1834. Ubersicht der Cyperaceen gattungen. Linnaea 9: 
286. 

, 1835-36. Cyperaceae Capensis secundum Novissimas Ecklonii collec- 

tiones. Linnaea 10: 139. 

Oliver, D., 1975. The botany of the Speke and Grant Expedition. Trans. Linn. Soc. 
Lond. 29: 166. 

Podlech, D., 1960. Uber Enige Cyperaceen Sudafrikas. Mitt. Staatssamml. Munch. 
3: 521-530. 

Ridley, H. N., 1884. West African Cyperaceae in the Welwitsch Herbarium. Trans. 

Linn. Soc. Lond. ser. 2, 2: 145-147. 

Robinson, E. A., 1959 (unpublished). Kyllinga in Tropical Africa. 

RottbPll, C. F., 1773. Descr. Ic. Nov. PL p. 13 t. 4f.4. Copenhagen. 

Schonland, S., 1922. Introduction to South African Cyperaceae. Mem. bot. Surv. S. 
Afr. 3: 19-21. Plates 1-3. 

Stafleu, F. A., 1967. Taxonomic Literature. Zug, Switzerland: I.D.C. 

(ed.), 1978. I.C.B.N. Utrecht: Bohn, Scheltema & Holkema. 

Vahl, M., 1806. Enumeratio Plantarum. 2: 382. Copenhagen. 


J1 S. Afr. Bot. 49 (3): 305-309 (1983) 


TYLECODON KRITZINGERI: A NEW TYLECODON (CRASSULACEAE) 
FROM THE NORTH WESTERN CAPE 

E. J. van Jaarsveld 

( National Botanic Gardens of South Africa , Kirstenbosch, Private Bag X7, 
Claremont 7735, R.S.A.) 


Abstract 

A new species of Tylecodon (Crassulaceae) is described from the north western 
Cape. 

UlTTREKSEL 

TYLECODON KRITZINGERI: 'N NUWE TYLECODON-SOORT VANAF 
NOORDWES-KAAP 

’n Nuwe Tylecodon-soort (Crassulaceae) vanaf Noordwes-Kaap word beskryf. 

Key words: Tylecodon , sp. nov., Crassulaceae, north western Cape. 

Tylecodon kritzingeri E. J. van Jaarsveld, sp. nov; a Tylecodon suffultus 
Bruyns ex Toelken differt, multo maiore plante, erecta ad 1 000 mm alta. 

Plantae perennes, ramis scandentibus gracilibus, 3-4 mm in diametro, 
radicibus tuberosis saepe ramosis. Folia linear-elliptica. Inflorescentia 
thyrsus, 1-3 monochasiis, quoque 1-3 floribus. Sepala triangulari-lanceo- 
lata 5 mm longa, acuta, glabra, succulenta, viridia. Corolla infundibularis; 
tubo 20-25 mm longo glabro; lobi 8-10 mm longi, lanceolati, acuti, glabri, 
sanguinei; stamina 12 mm longa, filamentis rectis, pilis patentibus praecipue 
in partibus quibusconnatis tubo petalorum. Carpella ovariis gracilibus grada- 
tim constructis in stylos erectos stigmatibus terminalibus. 

Typus: Cape Province — 2817 (Tatasberg): Lower N W slope of Oemsberg (-AC), 
Van Jaarsveld and Kritzinger 6278 (NBG, holotypus). 

Plants perennial, glabrous in all parts (except the filaments) single- 
stemmed or rarely branched, scrambling or scandent into bushes up to 1 m. 
Roots tuberous and branched, up to 10 mm in diameter. Stems slender, light 
green, 3-4 mm in diameter and succulent at first, erect and later spreading, 
becoming grey with dark striations, woody and snapping when bent. Leaves 
alternate and spirally arranged, linear to linear-elliptic and acute at apex, 


Accepted for publication 30th March, 1983. 

305 


306 


Journal of South African Botany 


erectly spreading and often becoming recoiled acting as tendrils, grooved on 
the adaxial surface, abaxial surface convex, becoming deciduous towards 
summer. Inflorescence a spreading thyrse up to 200 mm in diameter with 1-3 
monochasia each bearing 1-3 spreading flowers. Calyx glabrous, lobes tri- 
angular-lanceolate, and acute, 5 mm long and 2 mm broad. Corolla pale 
green, 20 mm long, 4 mm wide at the base and gradually expanding to 7 
mm, lobes 8-10 mm long, lanceolate and acute at apex, at first erect and be- 
coming erectly spreading; inner surface maroon to dark maroon, lower half 
with green striations. Filaments light green, 12 mm long, fused to corolla in 
lower half and sparingly pilose respectively, squamae yellowish-green, quad- 
rangular and emarginate, 1 mm x 1 mm. Carpels with slender ovaries grad- 
ually tapering to erect styles. Stigmas terminal. 

Flowering period: January. 

Distribution and Habitat 

This species occurs in the upper Gannakouriep river valley on the north 
west slopes of the Oemsberg, Western Mountain Karoo (Acocks Veld type 
no. 28), where it is scandent among xerophytic shrubs. The plants are locally 
common. Rainfall occurs mainly in winter and ranges from 150-200 mm per 
annum. 

Discussion 

Tylecodon kritzingeri van Jaarsveld is a very distinct species and shares, 
with Tylecodon suffultus Bruyns ex Toelken, the habit of being the only 
scrambling members of the genus. However, the two are not closely allied 
and are quite distinct. 

Tylecodon kritzingeri van Jaarsveld is a much larger plant with large fun- 
nel-shaped glabrous maroon flowers which bear erectly spreading lobes. 
(Tylecodon suffultus Bruyns ex Toelken is as yet only recorded from the 
Vanrhynsdorp district and is a much smaller plant with thicker branches and 
pink flowers; the corolla also possesses glandular hairs on the outside). The 
linear leaves of T. kritzingeri which often recoil and act as tendrils are 
unique in the genus. This species is the fastest growing in the genus. Plants 
that were grown under controlled glasshouse conditions at Kirstenbosch pro- 
duced stems 700 mm tall within one growing season. The open corolla per- 
sists for approximately 8 days. 

Tylecodon kritzingeri was first discovered on the 21st September 1981 by 
the author on a collecting expedition together with Mr. Kobus Kritzinger of 
the Cape Department of Nature Conservation. I have pleasure in naming 
this species after Mr. Kritzinger who has made a substantial contribution to 
nature conservation in Namaqualand and particularly in the Richtersveld. 



Distribution of Tylecodon kritzingeri. 



Tylecodon kritzingeri: a new Tylecodon 


309 


Mr. Kritzinger is at present Officer-in-charge of the Hester Malan Nature 
Reserve at Springbok and has collected many interesting records of plants in 
that area and has also taken the author to many inaccessible areas. 

Acknowledgements 

Mrs. E. Ward-Hilhorst is thanked for her beautiful illustration, and Mr. 
Norval Geldenhuys for checking the Latin diagnosis. 

References 

Tolken, H. R., 1978. New taxa and combinations in Cotyledon and allied genera. 
Bothalia 12 , 3: 377-393. 



J1 S. Afr. Bot. 49 (3): 311-322 (1983) 


Book Reviews 

The Monocotyledons: a Comparative Study, by Rolf M. T. Dahlgren and 
H. Trevor Clifford, with pp. 378. London: Academic Press, 1982. (“Botani- 
cal Systematics”: an occasional series of monographs, edited by V. H Hey- 
wood). US$ 98,00, £48,00. ISBN 0-12-200680-1. 

The familial and ordinal classification of the flowering plants has always been a 
complex problem fraught with the difficult questions of homology, convergence and 
character significance. Nevertheless the search for better and more natural schemes 
continues. And with time more and increasingly diverse kinds of data have become 
available to be applied to the problems of classification. At the same time the dif- 
ficulties of assembling, assimilating and interpreting the great quantities of data are 
compounded. Rolf Dahlgren and Trevor Clifford have produced in this book a re- 
markable effort at presenting much of the available data that can be brought to bear 
on systematic problems for the monocotyledons and synthesizing it into a coherent 
whole in the form of a revised scheme of classification for the monocots. 

Data of systematic value have been brought together from innumerable sources 
and condensed into a form that the reader can digest. The larger part of this impres- 
sive work is a survey of the “distribution of selected characters and their states”. 
Some 100 characters are dealt with, most of them accompanied by appropriate illus- 
trations, and their occurrence is plotted on bubble diagrams of those orders of the 
monocots which the authors recognize in their classification scheme, presented in an 
earlier chapter. In these diagrams, introduced and used by Dahlgren in a series of 
publications on plant phylogeny and classification, the bubbles represent the orders 
of monocots arranged spatially to indicate in a simple way, broad evolutionary re- 
lationships. 

A wide variety of characters are dealt with including traditional organ morphol- 
ogy (e.g. ovary position, ptyxis, venation patterns); as well as anatomy (e.g. latici- 
fers, stomata structure and origin, vessel shape); embryology (e.g. pollen morphol- 
ogy, tapetum types, ovule structure, placentation); chemistry (e.g. steroid saponons, 
cyanogenic compounds, flavonols and flavones); and a selection of less unorthodox 
features such as fine structure (sieve tube plastid bodies); and host specificity (e.g. 
plant lice, various fungi). 

A fundamental paradox, common to systematic studies in general, is that a sys- 
tem of classification is developed by the evaluation and synthesis of data, and then 
the classification is used to present and discuss this data. Although this may seem self 
serving, it is the most efficient way to present and deal with systematic situations. To 
be fair, the authors present a thorough historic review of monocot classification be- 
fore developing their own. One of the outstanding features of the Dahlgren-Clifford 
classification is their treatment of the order “Liliales” in its current broad sense, and 
it is this aspect which will have most impact on southern African systematists. This is 
broken down into several orders, the fundamental division being the separation of 
Asparagales from Liliales as well as the recognition of Burmanniales, Orchidales and 
isolated small orders like Haemodorales, Velloziales, Philydrales, and Bromeliales. 
Within Asparagales, united especially by the frequent presence of a black, phyto- 
melan seed crust, are several families not often accorded this status including some 
important in Africa like Asparagaceae; Dracaenaceae; Amaryllidaceae (here not in 
eluding Alliaceae); Tecophilaeaceae; Hyacinthaceae (Scilloideae), and Asp o 


311 


312 


Journal of South African Botany 


delaceae (Anthericoideae). Apart from a now relatively small North Temperate Li- 
liaceae (14 genera), Liliales comprises Iridaceae (the largest), Colchicaceae 
(Wurmbaeaoideae) and some non-African families, importantly Melanthiaceae, also 
largely North Temperate. 

Many of the families accorded recognition here have most often been included in 
a large, very diverse and heterogeneous “Liliaceae” in the past, sharing little except 
a generalized petaloid monocot ground plan. The system presented here is one of 
several developed in the last few years and it will take time for a consensus to evolve 
on which is the most useful, practical and correct (if any one system fulfils all these 
requirements). The Dahlgren-Clifford system seems to be for the most part, a natu- 
ral scheme and the most acceptable one so far developed and it is certainly the most 
fully documented classification for the monocots yet presented. Despite this, it will 
take many years before the system or a modification of it will be accepted. Such is 
the conservative and traditional bias in plant systematics. 

An interesting section near the end deals with the vexing question of the relation- 
ship of the monocots to the dicotyledons. There are several current hypotheses deal- 
ing with the problem, all reviewed by Dahlgren and Clifford, who favour the idea 
that Dioscoreales are close to the basic monocot stock and that this group has its 
closest affinities with Aristolochiales. This in turn is close to the basal magnolioid 
lines that comprise the most primitive of the living dicots. This runs counter to the 
more widely favoured theory that the aquatic monocots, Alismatales sens. lat. are 
the most primitive monocots and that they are allied to the dicots through the Nym- 
pheales. 

This book is a model of its kind in plant systematics and no botanical library can 
afford to be without it. It will be a valuable teaching aid, a useful reference for senior 
students as well as a fine research resource for systematists. 


Peter Goldblatt 

Plant Carbohydrates I: Intracellular Carbohydrates, edited by F. A. 
Loewus and W. Tanner, with pp. xxii + 918 and 103 figures. Berlin, Heidel- 
berg, New York: Springer-Verlag, 1982. Cloth, approx. US $132,40. ISBN 
3-540-11060-7. 

This 1 840 g work represents volume 13A of the New Series of the well known and 
much appreciated “Encyclopedia of Plant Physiology”. Together with Plant Carbo- 
hydrates II, it replaces the single volume of the original “Encyclopedia of Plant 
Physiology” which in 1958, adequately summarized the state of our knowledge about 
plant carbohydrates at that time. 

This volume provides a comprehensive assessment of our current knowledge of 
the carbohydrates which usually do not occur outside the plasmalemma of living 
plant cells. For convenience, all organisms except animals and bacteria are regarded 
as plants and the treatise deals with the carbohydrates of fungi and algae in addition 
to those occurring in mosses, ferns and seed plants. 

In contrast to what some might have expected, the emphasis throughout the 
book is on the physiological activity and metabolic involvement of the various carbo- 
hydrates rather than on their chemistry — as indeed befits a volume of an “Encyclo- 
pedia of Plant Physiology”. 

The 21 chapters which have been written by 62 well-known experts, are arranged 
in three sections. The first deals with the occurrence, metabolism and functions of 
the monosaccharides and oligosaccharides and related compounds. Special treatment 


Book Reviews 


313 


is given to sucrose and related oligosaccharides, as well as to the uronic acids, amino 
sugars, branched-chain sugars, sugar alcohols, cyclitols and heterosides (glycosides). 

The second section deals with the occurrence, metabolism and function of those 
macromolecular carbohydrates which occur intracellularly, i.e. starch and other re- 
serve polysaccharides, glycoproteins, glycolipids and steryl glycosides. Because of the 
importance of glycoproteins as lectins and as constituents of plasmamembranes, two 
chapters are devoted to this important class of compounds. 

In the third (final) section five chapters are devoted to purely physiological as- 
pects of carbohydrates: the translocation of sugars; the secretion of nectar; the stor- 
age of sugars in higher plants; the storage of starch; and the control by external and 
internal factors over the mobilization of reserve carbohydrates in higher plants. 

The recommendations of the International Union of Pure and Applied Chemistry 
(IUPAC) regarding carbohydrate nomenclature, the rules of the International Union 
of Biochemistry (IUB) on oligosaccharides and polysaccharides and the recommen- 
dations of the IUPAC-IUB Commission on Biochemical Nomenclature were ad- 
hered to throughout. 

This monumental work will no doubt become a standard reference work on plant 
carbohydrates for several years. It is clearly written and well illustrated. The many 
references at the end of each chapter are especially useful because the full titles of 
the papers and the lengths of the papers cited are provided in all cases. Apart from 
the usual subject index, an author index and a species index are also provided. 

N. Grobbelaar 


Plant Carbohydrates II: Extracellular Carbohydrates, edited by 
W. Tanner and F. A. Loewus, with pp. xxii + 769 and 124 figures. Berlin, 
Heidelberg, New York: Springer-Verlag, 1981. Cloth, approx. US $124,80. 
ISBN 3-540-11007-0. 


This book represents volume 13B of the New Series of the authoritative Encyclo- 
pedia of Plant Physiology" which has become a standard reference work on plant 
physiology. It deals with those carbohydrates and related compounds which are se- 
creted by the protoplasm of plant cells and therefore usually occur outside the plas- 
malemma of such cells. 

The carbohydrate components of the cell walls of fungi, algae and seed plants are 
dealt with in considerable detail. This is necessitated by the relatively recent discov- 
ery of the importance of carbohydrates in cell-cell surface interactions and recogni- 
tion systems. Thus host-parasite and pollen-pistil interactions as well as symbiosis 
and the mating reaction in fungi are but some of the complex but fascinating p ysio 
logical phenomena in w'hich extracellular plant carbohydrates play key roles. 

The book contains 26 chapters which have been written by 38 recognised experts 
in their respective fields. The chapters are grouped into five sections and the nrst, 
which deals with the cell walls of higher plants, contains no less than c ap e . 
They cover the constitution of the cell wall polysaccharides; the biophysica as w 
the biochemical ultrastructure of the cell wall; the assembly of polysaccharide fibrils, 
the biosynthesis and metabolism of cellulose and noncellulosic cell wa P Y 
rides; the glycoproteins and enzymes of the cell wall; the role of hpi - m e 
ides in the biosynthesis of complex carbohydrates; the biosynthesis o ig - ’ 

suberin and associated waxes; and finally the effect of hormones on ce 


314 


Journal of South African Botany 


Section two is devoted to the cell walls of algae and fungi — their composition, 
biosynthesis, cytology and degrading enzymes (autolysines) — eight chapters. The 
three chapters of section three all deal with secretion — especially as it occurs in 
fungi, the root cap of seed plants and in cell wall formation during cell division. 

Sections four and five are probably the most topical. They deal with cell wall sur- 
face phenomena including lectin-carbohydrate interactions (2 chapters each). The in- 
volvement of carbohydrates in pollination and plant-pathogen interactions is dis- 
cussed in section four. In the last section the physiological role of lectins in slime 
moulds and higher plants and in symbiotic plant-microbe interactions are discussed. 

The presentation of the material, although encyclopedic, is excellent. It reads 
easily and is quite lucid and well illustrated. The references at the end of each chap- 
ter provide the full title and length of each publication. Apart from the subject in- 
dex, an author index is also provided. A species index is, however, not included. 

Because this volume deals with the more novel aspects of plant carbohydrate 
physiology, it is bound to become outdated sooner than Volume I. It is nevertheless 
a must for each University Botany Department and plant physiological research es- 
tablishment and will no doubt be used as a standard reference work for several years 
to come. 


N. Grobbelaar 

Growth and Differentiation in Plants, by P. F. Wareing and I. D. J. 
Phillips, with pp. xi + 343. Oxford, New York, Toronto, Sydney, Paris, 
Frankfurt: Pergamon Press, 1981. US $20 (F), $40 (H). ISBN (F) 
0-08-0263-510. 

The third edition of Wareing’s and Phillips’ plant physiology textbook shows few 
differences from the second. The title has been shortened, some references added, 
the chapters organized into four sections, and each section preceded by a brief intro- 
duction. The chapter sequence, however, has been maintained. The heading of the 
last chapter has been changed from “The Control of Development” to “Gene Ex- 
pression and Cell Determination in Development” and in the process the treatment 
of polarity moved to the first chapter “The Plant Cell in Development”. New infor- 
mation added to this chapter includes that on the cell cycle and secondary wall for- 
mation. 

The third edition has undergone a 25% increase in surface size without a corre- 
sponding increase in information, unfortunately making it a less wieldy book than its 
predecessor. The micrographs, generally, have been poorly reproduced (examples: 
Fig. 1.10, p. 12; Fig. 2.3, p. 24); some are clearer than the corresponding ones in the 
2nd edition, but others are poorer by comparison (example: Fig. 2.5 B — designated b 
on the photograph — p. 25). Some graphic reprintings, compared with the original 
publications, have lost definition (examples: Fig. 9.1, p. 218; Fig. 11.1, p. 261); many 
line drawings and chemical configurations lack boldness (examples: Fig. 2.18, p. 38; 
Fig. 4.3, p. 79) and on some the lettering and symbols are too small and faint (ex- 
ample: Fig. 5.31, p. 140). There is, furthermore, much wasted space (some ex- 
amples: pp. 32, 103, 116, 130, 166); space which could for example have been util- 
ized to increase magnification of photographs, graphs and line drawings. As 
compared with an American or German botanical textbook I would have to judge 
the general design and layout as inferior. 

Despite the overall inexcellence of the book’s production, there is much to com- 
mend it. Dealing only with growth and development, it is not a general plant physiol- 


Book Reviews 


ogy text, but one that should be able to treat this complex discipline with authority. 
This, as is to be expected from Aberystwyth, certainly is the case in respect of the 
two central sections covering internal and environmental controls of development. 
The chapter on growth movements, in particular, has been expanded and greatly im- 
proved. The book is extremely readable, as up-to-date as one can expect a book of 
this nature to be (although I found no reference to Skoog’s Plant Growth Substances , 
1979), and displays a fresh perspective by relating phenomena under discussion to 
the experiments used in their investigations. It is a good teaching text. 

To soften the comments made above about the illustrations: a reviewer in Nature 
(1982) pointed out the transformation which the flowering test-tube Primula under- 
went on the 3rd edition's cover page. The artist certainly has made it represent plant 
growth and development much more colourfully than was the case with its predeces- 
sor on the cover of the 2nd. 


Chris H. Bornman 

Experiments in Plant Tissue Culture, by J. H. Dodds and L. W. Roberts, 
with xiii 4- 178. Cambridge, London, New York, New Rochelle, Mel- 
bourne, Sydney: Cambridge University Press, 1982. £7,50. ISBN 

0-521-23477-8 (hard cover); 0-521-29965-9 (paperback). 

Although the number of books on plant tissue culture has increased dramatically 
over the past decade, few have dealt explicitly with methodology. One of the first 
was R. A. de Fossard’s excellent, inexpensive and comprehensive Tissue Culture for 
Plant Propagators (The University of New England Printery, Armidale, Australia, 
1976), and another the expensive, far less comprehensive Plant Tissue Culture: A 
Laboratory Manual* by Reinert and Yeoman (Springer-Verlag, 1982). Now there 
has appeared (1982) a book that is packed with information and one that takes an 
approach which is different compared with the firstmentioned two. 

The book under review comprises 14 chapters and is furnished with author and 
subject indexes, as well as lists of terms and commercial sources of supplies. Chapter 
1 gives an interesting account of the history of plant tissue culture up to the present 
day. Chapters 2 and 3 treat the problems of aseptic techniques and culture media, re- 
pectively. Chapters 4 to 13 involve experiments that demonstrate the use of plant or- 
gan, tissue, cell and protoplast cultures as research tools. The final chapter is on a 
much neglected aspect of plant tissue culture, namely quantitation of procedures. 

A useful feature of the book is the organization of each chapter. It begins with a 
brief historical background, the development of the problem, and a statement of pur- 
pose; then comes a list of materials required, a description of the procedure and re- 
sults, and a number of questions for discussion; and finally, there follows an appen- 
dix with additional experiments and techniques, and a list of selected references. 

It is obvious that the authors (JHD — a young, gifted and progressive plant bi- 
ologist and LWR — well-known and respected for his book Cytodifferentiation in 
Plants. Xylogenesis as a Model System: Cambridge University Press, 1976) with this 
book will reach an audience that includes researchers from a variety of botanical dis- 
ciplines, in addition to the undergraduate and advanced level high school students 
for whom it was intended. It goes further than being merely a laboratory textbook, it 
is also a useful source to most of the major studies that have high-lighted the devel- 
opment of plant tissue culture. 


Reviewed on page 317 of this issue. 


316 


Journal of South African Botany 


Some errors, not all typographical ( Haplopappus gracilus instead of gracilis in 
the index), have crept into the book. Examples: The generic name of tomato is spelt 
Lycopersicum , instead of Lycopersicon , throughout the book (also in the reference 
to Padmanabhan, p. 87). Sugar beet. Beta vulgaris is said (p. 124) to have a tuber, 
instead of a fleshy hypocytol-root. The major shortcoming of the book is to be found 
in chapter 10 on the isolation, purification and culture of protoplasts. Reference 
could have been made here to the eminently more elegant techniques of J. F. Shep- 
ard and colleagues. Also, the dismissal of plant regeneration from protoplasts with a 
single short paragraph will certainly convey the false impression that it is but a few 
easy steps, when in fact it is very problematical indeed. Further, it surely is also Wid- 
holm (1972) and not only Larkin (1976) who deserves credit for assessment of viabil- 
ity using fluorescein diacetate. 

Experiments in Plant Tissue Culture is a well-produced and highly informative 
textbook and one that is destined to be widely used. 


Chris H. Bornman 


Plant Physiology, by I. P. Ting, with pp. ix + 642. Reading (Massachusetts), 
Menlo Park (California), London, Amsterdam, Don Mills (Ontario), Sydney: 
Addison-Wesley Publishing Company, 1982. US $27,95. ISBN 0 201 07406 0. 

Yet another relatively high-priced plant physiology text! What novel features 
might it offer? This question is not easily answered. 

Much attention has been devoted to design. The subject matter of the 21 chap- 
ters has been divided variously over five parts. Each part's title is announced separa- 
tely on a page in enormous headlines, below which there are what is referred to as 
“chapter opening photographs”, one photograph for each chapter that follows. Part 
II, with six chapters, has six slim chapter opening photographs squeezed alongside 
each other. Each photograph, but now magnified 3x, reappears again with the intro- 
duction to the chapter. Following the page that announces the part, is a one or two 
page “prologue” which serves as a summarized introduction of the chapters compris- 
ing the part. The part is concluded by a “prospectus”, a page or two that briefly re- 
views hypotheses, indicates the current state of research, and speculates briefly on 
future trends. The chapter opening photographs in reality have little if any bearing 
on the chapters, and are obviously intended to attract and hold the reader’s atten- 
tion, as are the bold black lines that run across the pages underlining chapter sub- 
headings. These features, unfortunately, detract from rather than add to the quality 
of the book. 

The five parts are: 

(1) Structure and function of cells and tissues. This is an elementary introduction 
of plant cell (and tissue) structure and function which follows a prologue in 
which the author expresses the view that the concept of “structure-function is 
somewhat dangerous because it tends to lead to teleological interpretation”. 

It is understandable that liberal use was made of light and electron micrographs 
supplied by colleagues from institutions on the same campus (UC Riverside). How- 
ever, a number of the photographs by no means represent the best illustrative ma- 
terial available (example: Fig. Ell, vacuole-extruding protoplasts). The information 
on plant classification could better have been omitted. The author refers to the 
groups, divisions, orders and classes listed in Tab. 1.2 as species. 

(2) Biophysical processes: exchanges with the physical environment, (3) Biochem- 
ical constituents of plants and (4) Metabolic processes of plants. These three 


Book Reviews 


317 


parts form the bulk and heart of the book, making up for lack of depth by 
the breadth of their coverage. 3 

(5) Growth and development: integrative processes. Like part 1, this final section 
is elementary. It can at best be regarded as a quick, simplified scan of a disci- 
pline which the author (in the prospectus) regards as the most complex of 
plant physiology. 

In particular, growth movements, the concepts of morphogenesis, differentiation 
and determination, and the plant hormones and growth regulators have been accord- 
ed shallow treatment. 

It is difficult to judge to whom the book is addressed. If it is the undergraduate, 
it would seem that an introductory text such as Biology of Plants (Raven, Evert and 
Curtis: Worth) would be a superior choice, for in addition to physiology it also em- 
braces the other disciplines of botany. If it is the graduate, there are more meticulous 
(Lehrbuch der Pflanzenphysiologie — Mohr and Schopfer: Springer), comprehensive 
(Plant Physiology — Salisbury and Ross: Wadsworth), scholarly ‘'(Plant Physiology— 
Thomas, Ranson and Richardson: Longman) and readable (Plant Physiologv— Bid- 
well: MacMillan) textbooks available. 


Chris H. Bornman 


Plant Cell and Tissue Culture: A Laboratory Manual, by J. Reinert 
and M. M. Yeoman, with pp. 83. Berlin* Heidelberg, New York: Springer- 
Verlag, 1982. US$22. ISBN 3 540 11316 9. 

As indicated in the title, this book is a laboratory manual. It is aimed at students 
and others who are inexperienced in the techniques of plant tissue culture. The 
authors state clearly that there was no pretence of sophistication and that the experi- 
ments can be carried out with a minimum of facilities. 

Altogether 17 experiments have been selected, from the simple establishment of 
callus to the more complex isolation and culture of leaf protoplasts, and arranged in 
six groups. The first set (Part II) of four deal with excision and explantation of tissues 
and the initiation of cultures, that is cell division and growth. Still dealing with cell 
division and growth. Part III includes two bioassay experiments for kinetin. Part IV 
covers morphogenesis and its four experiments illustrate plant regeneration in vitro 
via somatic embryogenesis from cell cultures and immature microgametophytes, 
from callus tissues, as well as from pseudobulbs. In Part V there are two experiments 
on the isolation, culture and fusion of protoplasts, and two in Part VI on secondary 
product synthesis. The last part (VII) includes three experiments on embryo and or- 
gan culture. 

Obviously, scores of additional experiments covering a wider range of agronom- 
ic, ornamental, floral and vegetable plants could have been included; or, alternative- 
ly, one or two experiments demonstrating epigenesis (habituation, for example) or 
mutational changes might have added some zest and challenge to the objectives. 
However, since the authors made their intent quite clear in their introductory re- 
marks, one should be grateful that a few plants other than the two classic, over- 
worked models of carrot and tobacco also feature in this volume’s contents. 

The experiments are presented clearly and concisely and are all relatively easy to 
carry out. However, the methods described for the isolation of mesophyll protoplasts 
are no longer used by anyone seriously interested in protoplast isolation, culture an 
plant regeneration, as they involve some time-consuming operations as well as sedi- 
mentation of the protoplasts by centrifugation. Missing, too, is information on tie 


318 


Journal of South African Botany 


use of stains and other methods to ascertain not only the viability of the protoplasts, 
but also the presence of subprotoplasts (cytoplasts, vacuoplasts, microplasts) so that 
plating densities can be compensated for. 

The layout uses an interesting approach even though it is reminiscent of that 
found in some modern cookbooks. The items required for an experiment are dia- 
grammatically depicted in a box titled: “Items for the sterile transfer room”. The 
items themselves are numbered, the numbers corresponding to the same listed under 
Materials and Equipment. So, for example, if a scalpel is listed sixth, it is found 
boxed in below and identified with the number 6. However, in succeeding experi- 
ments this item may have the number 5 or 7 or 11, and so forth. It would in my opin- 
ion have been better to present the layout in the transfer room only once, with all 
the items for all experiments boxed-in and numbered. The number would then have 
served the purpose also of permanent identification and the space of 18 boxes (ca. 
2 500 cm 2 ) would have been conserved. As it is, 16 of the book’s 83 pages are virtual- 
ly blank, making it, at US $0,33 per page, a very expensive book indeed. Compare 
this with the average price of US $0,11 per page of technical books in the natural sci- 
ences reviewed in Science (79: 47, 1983) for the year 1982. As far as being aimed at 
the inexperienced student, it can be reasonably predicted that this manual is destined 
for the copying machine. 

(The book suffered the unfortunate fate of being recalled by the publisher, after 
it was discovered that the Index was incorrect.) 


Chris H. Bornman 


Saline and Sodic Soils, by E. Bresler, B. L. McNeal and D. L. Carter, 
with pp. x + 236 and 78 figures. Berlin, Heidelberg, New York: Springer- 
Verlag, 1982. Cloth DM 98, approx. US $43,60. ISBN 3-540-11120-4. 

This book is another in the Springer-Verlag “Advanced Series in Agricultural 
Sciences”, and conforms to the general format and quality of the series. It is well 
written with very few errors, and is usefully illustrated. 

The book is in three approximately equal parts: 1. Diagnosis and Properties; 
2. Transportation and Distribution of Salts; 3. Management. 

Diagnosis and Properties deals with the sources of salts, some water quality and 
soil solution parameters, surface phenomena of salt-affected soils, salt dissolution 
and precipitation and the diagnosis of salinity properties. In general this section is 
comprehensive and well written, though more attention could have been given to the 
origins of sodic soils. A sample of saline and sodic soil forming minerals are men- 
tioned but, for example, the sodic soils in granite areas where the granite contains 
albite (common in south-central Africa) are not included. The theoretical treatment 
of the exchange equilibria and other aspects of deposition, etc., is detailed and well 
illustrated. 

The section on transportation and distribution of salts is comprehensive but will 
be difficult to follow for the average biologist who is not very numerate. The sixty 
pages of Section 2.3 (Modelling of Salt Flow Phenomena) are mostly beyond the 
needs and comprehension of botanists. 

The final 45 pages on Management are interesting and of practical value, and 
written in a simple, easily understood style. Tables 15-18, which detail the relative 
productivities and salinity thresholds of a number of crop types, from sensitive to tol- 
erant, under a range of increasing soil salinities, is a particularly useful reference. So, 
too, are the final sections on reclamation and specialized management practices. 


Book Reviews 


319 


I doubt that the book will become part of the average botanist’s personal library, 
but I strongly recommend its inclusion in all general libraries. Although the examples 
are mostly American, the general principles apply universally and it will be an im- 
portant reference and text for research workers in soil science, crop production and 
general plant/soil relationships. 


B. H. Walker 

Experimental Embryology of Vascular Plants, edited by B. M. Johri, 
with pp. 273 and 81 figures. Berlin, Heidelberg, New York: Springer-Ver- 
lag, 1982. DM 98, approx. US $43,60. ISBN 3-540-10334-1. 

During the past decade a large number of books have been published which dealt 
with the general field of tissue culture. In view of this, it is only natural to question 
whether the publication of the current volume was necessary? In view of the rapid 
advances which have been made in the field of experimental botany this book is en- 
tirely justified, more so, as it is intended as a brief and concise account of some of 
the more recent developments in this rapidly changing field. The book is aimed to- 
wards the undergraduate and postgraduate student. As such it was emphasized by 
the editor that no exhaustive literature reviews have been included. This does not de- 
tract from the various chapters which in the reviewer’s view contain more than suf- 
ficient literature to stimulate the curiosity of young prospective students. 

The book consists of a general introduction and nine other chapters which cover 
a wide spectrum of topics. They include: experimental embryology of pteridophytes; 
experimental embryology of gymnosperms; flower culture; anther culture; ovary, 
ovule, and nucellus culture; pollen-pistil interaction and control of fertilization; en- 
dosperm culture; embryo culture and protoplast culture. All chapters give a brief re- 
view of the literature and then concentrate on some of the more recent discoveries 
and developments in the respective fields under discussion. Throughout a fairly criti- 
cal approach has been adopted. Attempts were made to give impetus to future re- 
search by the identification of some of the problems which have been encountered in 
some fields of study. 

The volume is neatly presented and the printing and figures are of a high quality. 
The book is reasonably priced and should be on the shelves of all teaching and re- 
search institutions. 


J. van Staden 


Historical Plant Geography, by Philip Stott, with pp. 151. London: 
George Allen & Unwin, 1981. R16.60. ISBN 0-04-580010-3; 

0-04-580011-1 (pbk.) 

The saying that a geographer has to be a scientific Jack-of-all-trades finds its confir- 
mation in this concise book of 151 pages. Mr. Stott has brought a complicated branc 
of botanical science in all its ramifications to the student as well as to the intereste 
layman. The book starts with a discussion on the suitable selection of study areas 
whether limited to species or genus levels or a continent-wide family leve ( ap 
ter 1). In Chapter 2 the importance of correct taxonomy, based on reliable annotated 
collections and well-prepared herbarium material, is duly stressed. The discussion o 
the term “type” on page 21, however, could have been clearer. The last section on 


320 


Journal of South African Botany 


advanced methods should perhaps have been transferred to Chapter 3 which dwells 
on the methods applied in the production of plant maps and the selection of the best 
suitable projections. Although great importance is laid on a show of the selected pro- 
jection and of a bar scale, some of the map examples should also have shown at least 
one latitude and longitude, especially on maps of remote regions, as for instance 
figure 3.6 on page 40. 

Chapter 4 deals with patterns of distribution based on the floristic elements of the 
various distribution types in the world. The use of the word territory on page 47 is 
somewhat misleading as in zoology the terms denotes an area which is defended by 
an animal against others of the same species. Chapter 5 investigates the reasons be- 
hind the various distribution patterns. The author delves into the multiplicity of a 
number of ecological phenomena to arrive at conclusions on the origin of distribu- 
tions and their boundaries, based essentially on the autecology of the species in ques- 
tion. In this connection he quotes Holland’s “An evolutionary biography of the gen- 
us Aloe" (J. Biogeog. 5, 213-226) as an attempt to find the centre of origin for the 
genus. In principle this may be sound enough, although I feel that the taxonomy of 
this genus is in need of revision, particularly that of the sessile and shortstemmed 
maculate group where systematic sampling would almost certainly reveal clinal phe- 
nomena with consequent reduction in the number of species, especially in the epi- 
centre of distribution in the Transvaal. 

Well represented is the section dealing with the disruptions of distribution by the 
drift of the continents in the wake of plate tectonics and sea floor spreading. A few 
more examples of species affected by these geological events would have been wel- 
come. The section on climatic changes in the past is a little too brief. However, the 
train of thought is continued in Chapter 6 where the full importance of macro- and 
microfossils, especially that of pollen analysis, is emphasised. 

Chapters 7 and 8 on disjunct distribution patterns and the endemism problem are 
very clear and well represented, although some of the evidence, for instance plate 
7,1 dealing with the distribution history of the Proteaceae which is linked to plate 
tectonic events, could already have been brought in in Chapter 5. Later the problems 
of vicariousness and speciation under isolation are discussed. Island and high moun- 
tain populations are explained. In Chapter 9 the author dwells on the future influ- 
ence of cyto- and chemotaxonomy on plant geographical methods. Although these 
disciplines are becoming increasingly in the mode, it should be emphasised here that 
their techniques are to be considered more as an aid to taxonomy and systematics 
rather than the end product. If too vigorously applied, their final effect on taxonomi- 
cal terms such as species, subspecies and varieties by the creation of a multitude of 
hybrids might be rather disconcerting and impractical if taxonomical keys designed 
for botanists in the field are based thereon. 

In conclusion, I would state that this is a very readable book couched in easily 
understandable scientific language and well illustrated by drawings, diagrams and 
tables. The cover, more suitable for a history of taxonomy, is somewhat unattractive. 
The layout is good, the price reasonable. All interested in this fascinating subject, 
whether students or laymen, should try to read this book 


N. H. G. Jacobsen 

New Perspectives in Wood Anatomy, edited by Pieter Baas, with pp. vi + 
252. The Hague: Martinus Nijhoff Publishers BV, 1982. Dfl. 130, approx. 
US $54,00. ISBN 90-247-2526-7. 

For over 50 years the International Association of Wood Anatomists has cham- 


Book Reviews 


321 


pioned the scientific study of wood structure. The society and its many distinguished 
members have made important contributions to investigations of wood anatomy. Es- 
pecially noteworthy among the achievements are the standardization of wood termin- 
ology, the publication of scientific articles on wood structure and function, the en- 
couragement for developing a worldwide network of repositories for vouchered 
wood samples and wood microslides, and the establishment of a spirit for free ex- 
change of specimens and knowledge around the world, both in basic and applied 
areas. Over most of its history, publications of IAWA have covered many subjects, 
but especially many systematic topics and regional surveys of wood, so on their 50th 
anniversary one might have expected a book summarizing these achievements. 

Instead, IAWA once again demonstrated leadership by producing New Perspec- 
tives in Wood Anatomy (P. Baas, ed.), a commemorative volume on the latest tech- 
niques and interesting avenues of research for wood analysis. The nine chapters are 
extended versions of invited papers presented at the 1981 anniversary gathering at 
the International Botanical Congress in Sydney, Australia by leading authorities on 
wood study. 

The book begins with a chapter on the early history of IAWA as reconstructed 
by W. L. Stern from publications, private correspondence, and personal sleuthing. 
Not only is this an excellent and lucid account of the actual events and controversies, 
but also here one gains a real flavour for the personalities of S. J. Record, L. Chalk, 
and B. J. Rendle, the founders of IAWA and the principal pioneers charting its early 
course. Such an account will be treasured by present and future anatomists, who can- 
not obtain this valuable historical perspective in our modern educational setting. 

Much wood research has asked the question, “Why are woods similar or differ- 
ent?” Naturally, workers have searched diligently for the answers, cautiously hoping 
that similarities might help solve taxonomic and phylogenetic problems. In Chapter 
two, P. Baas briefly narrates this search, beginning with the earliest anatomists, and 
the author presents a new perspective on Antoni van Leeuwenhoek, who has not 
been given adequate credit for his descriptions and illustrations of wood. 

As investigators studied more woods, they became cautiously optimistic that 
wood structure would help to classify taxa, but eventually they had to adopt the con- 
clusion that wood anatomy is no more dependable than many other features — good 
systematics must carefully evaluate wood features along with all other comparative 
evidence. Baas mentions the Baileyan (phylogenetic) trends in wood evolution only 
briefly because this is pre-empted by several excellent reviews; however, the author 
does emphasize the important but now well-known caveat that many trends are re- 
versible under certain conditions, a fact of which I. W. Bailey himself was aware and 
encouraged in the 1960’s. Most of the chapter is devoted to “ecological wood ana- 
tomy,” resurrected by S. Carlquist in the 1960’s. Here investigators attempt to 
understand why woods change, in relation to what ecological factors. Unfortunately, 
Baas, like others, presents no concrete answers, only narratives based on broad stu- 
dies using the correlation approach. Data can be assembled to suggest this trend or 
that one, but, in fact, the answers will come when anatomists analyse how plant phe- 
nology, whole plant physiology, and growth habit, not merely habitat, are linked to 
wood structure. This chapter could have been more lucidly written. 

In the next chapter, M. H. Zimmermann discusses one aspect of the relationships 
between wood structure and physiology, hydraulic conductance. Much of this infor- 
mation has been published before, but this is a useful review. However, the author 
has also made an important contribution by emphasizing the importance of studying 
vessel length, whereas wood descriptions based on sections and maceintions always 
emphasize vessel-element length. . „ 

Chapter four is a fascinating review by the skilful electron microscopists B. . 
Butterfield and B. A. Meylan of the very recent research on formation of perforation 


322 


Journal of South African Botany 


plates in vessels. Anatomists will appreciate the detailed discussions on hydrolysis of 
cell wall components and still marvel at the mystery that hydrolysis degrades some 
parts of the cell wall more than others. Even in this last part of the 20th century, we 
have not yet elucidated the final stages in vessel pore formation! 

The two following chapters are excellent accounts of the origin of the vascular 
tissues. In his characteristic style, P. R. Larson deliberately and cautiously traces the 
origin of procambium and eventual development of cambium. Although the author 
probably did not have to reiterate the early stages of procambium in embryos, he 
clearly and expertly showed how the transition from procambium to cambium is a 
continuum. Specialists on plant structure have recognized this for a long time, but 
the point must be thrust forward because beginning students are still accustomed to 
thinking that primary and secondary growth are discrete and definable. Next G. P. 
Berlyn reviews the most recent information on developmental and morphogenetic 
events in the cambial zone. So much needs to be learned about the impact of mor- 
phogenetic factors on the cambial initials and derivatives, and Berlyn’s chapter gives 
leads for future research. 

Every wood anatomist is well aware that wood samples can vary tremendously, 
depending on the place and climate of collection but also upon differences among in- 
dividuals in a single population or position within a single plant. These sources of 
variability are discussed by J. Burley with the goal to provide wood technologists 
with, knowledge for screening specimens in breeding programs. The techniques de- 
scribed should help workers identify genetically superior woods and save much time 
in trial-and-error research. 

In a similar vein, a chapter by Burley and R. B. Miller shows how statistics and 
computers are becoming indispensable in wood studies. In the future, much data- 
capturing on wood dimensions, including the storage and retrieval of measurements, 
the development of dichotomous keys, and the identification of unknown specimens, 
will be accomplished by computers. 

Perhaps the most difficult chapter to digest is by J. D. Boyd, which attempts to 
model the effects of physical stresses on layers of microfibrils in wood fibre cell walls. 
Unfortunately, most wood anatomists are not trained in biophysics, so the results 
will have a narrow audience. Nevertheless, the subject is especially important to 
wood technologists and wood breeders. Other models are needed to describe how 
physical stresses affect the wood characteristics and growth habits of living trees, so 
that we may eventually explain the numerous wood fibre designs in the secondary 
xylem of dicotyledons. 

Overall this volume is attractive and pleasant to read. Its subject is universally 
applicable to those doing research on wood. Especially laudable is the long list of lit- 
erature following each chapter. Unfortunately, the high price of this thin book will 
probably limit its distribution to libraries and to those researchers working directly in 
areas that are covered. 


Arthur C. Gibson 



J1 S. Afr. Bot. 49 (4): 323-327 (1983) 


WILHELM JAN LUTJEHARMS (1907-1983) 

On 3rd February 1983 South African Botany lost a true scholar and phil- 
osopher in the passing of Wilhelm Jan Liitjeharms, Emeritus Professor of 
Botany and Microbiology at the University of Cape Town. During his aca- 
demic career he had the distinction of occupying three individual chairs at 
South African Universities and his death will be particularly mourned by the 
many students, both English- and Afrikaans-speaking, who have benefited 
so richly from his wise and inspiring teaching at the Universities of the 
Orange Free State and Cape Town. 


Accepted for publication 16th May, 1983. 

323 

I 


324 


Journal of South African Botany 


Wilhelm Liitjeharms was born in Alkmaar, that little town some 35 km 
north of Amsterdam so famous for its cheeses, on 24 January 1907. It was 
there that he received his school education and also his early interest in mat- 
ters biological, inculcated during walks through the countryside with his 
grandfather, and by his collection of that remarkable series of scientifically 
accurate nature study albums produced by Jac. P. Thysse and the Verkade 
Company of Holland, which influenced so greatly an entire generation of 
Dutch youth. 

He entered Amsterdam University in 1924 to study for his Candidaat’s 
examinations in Natural Philosophy, majoring in Botany and Zoology. He 
graduated in 1927 and such was the standard of his academic achievements 
that, after a spell of lecturing at Amsterdam University, he was successful in 
gaining a post in the prestigious Rijksherbarium, Leiden; this in spite of his 
youth (Wilhelm was always the youngest student in his classes) and the great 
depression of that time which precluded most of his classmates from finding 
scientific employment. While working in the Rijksherbarium as Curator of 
the Mycological Collections, he was able to study further for his Doctoran- 
dus examinations under Professors Westerdijk (Plant Pathology), Weevers 
(Plant Physiology) and Stomps (General Botany), all leaders in their re- 
spective disciplines at the time. He then enrolled at the University of Leiden 
for his doctorate of philosophy under the direction of the famous scientific 
historian Professor Baas Becking. Even at this stage he had become interest- 
ed in South African Botany through his contact with the South African Col- 
lections of the Rijksherbarium and had originally intended to write his thesis 
on the life and works of Persoon, that famous South African Mycologist re- 
garded as the father of systematic mycology. In this aspiration he was, how- 
ever, frustrated as a senior German colleague had “adopted” Persoon for 
his own research (never published) and Wilhelm therefore switched his doc- 
toral project to a mammoth study of the history of Mycology during the 18th 
century. His brilliant thesis on this topic was published in the Mededeelin- 
gen van de Ned. Mycologische Vereeniging , deel 23, in 1936 under the title 
“Zur Geschichte der Mykologie DAS XVIII Jahrhundert”. During this 
period he married Anne-Marie Bloemendaal, a young lady from his home 
town whom he had known from school days and who was to prove such a 
staunch support to him throughout his life both as wife and academic col- 
league. 

As a result of his doctoral thesis, Wilhelm was awarded a Fellowship of 
the Royal Netherlands Academy of Science (Buitenzorg Fonds) for a six 
months study of the mycology of Java (the present Indonesia). Here he 
spent a happy and productive period with his wife, and was also able to 
carry out a botanical survey of the island of Enggano off the west coast of 
Sumatra for the Dutch Forestry Department and the Buitenzorg and Leiden 


325 


Herbaria. He came home with a whetted appetite for more exciting floras 
than that of his own native Netherlands. Thus, when the chair of Botany at 
the University College of the Orange Free State was advertised in 1938, he 
followed up his early interest in the flora of this country and applied for the 
post which, to his delight, he was successful in obtaining. 

He arrived at a very different University to the U.O.F.S. we know to- 
day. Money was scarce and very little apparatus of any sort existed in the 
department he had inherited from Professor Potts. Being somewhat disap- 
pointed by the scarcity of higher fungi in the Bloemfontein area he turned 
his attention to the soil micro-flora and so established the science of soil 
microbiology at the University. Improvisation was the order of the day. Au- 
toclaves were derived from pressure cookers; culture racks were made from 
paraffin tins cut in half and filled with sand. When he applied for a labora- 
tory assistant the University Council informed him that such a luxury was 
quite out of the question owing to shortness of funds, so his wife, Anne- 
Marie, became his unpaid assistant responsible for the subculturing of the 
fungal collection he was rapidly building up. Her remuneration was two cups 
of coffee a day. Gradually progress was made and, building on his earlier 
work in the Rijksherbarium, he was soon to establish himself as an interna- 
tional authority in the field of soil microbiology. 

During his professorship at Bloemfontein he travelled widely, thanks to 
the award of Carnegie, Rockefeller, C.S.I.R and Deutsche Forschungsge- 
meinschaft Fellowships, visiting England, Denmark, Sweden, Germany, 
Belgium, France, Canada and the United States to work with colleagues and 
to present papers at international congresses. He became the first professor 
to give lectures in Microbiology at any South African University and can 
therefore truly be called one of the founding fathers of this scientific disci- 
pline in the country. He was appointed to the Steering Committee for 
Microbiological Chemistry at the C.S.I.R., an office he held for many years 
(1961-1975), and served at a critical time in the country’s scientific develop- 
ment on a number of influential official committees including the Medical- 
Biology Committee of the Atomic Energy Board, the Committee on Inter- 
national and Scientific Co-operation and the Advisory Committee for 
Biological Sciences of the C.S.I.R. At his own university he was appointed 
Head of the Institute of Soil Biology which he had been so strongly influen- 
tial in founding. 

In 1962 Wilhelm Liitjeharms came to the Cape to occupy the Harry 
Bolus Chair of Botany at the University of Cape Town, the oldest botanical 
Chair in the country. At this time the University was striving to establish a 
microbiological component in its scientific curriculum and to W ; helm fell 
the task of co-ordinating these activities in addition to his already onerous 
duties as Head of the Botany Department. The success of his endeavours 


326 


Journal of South African Botany 


was witnessed in 1970 by the establishment within the University of a fully- 
fledged Department of Microbiology and his own appointment as first Pro- 
fessor of Microbiology, a post he held with distinction until his retirement in 
1972. 

Not only was Wilhelm a dedicated scientist but he had also a strong in- 
terest in the philosophy and history of Science. His papers on these subjects, 
in the writing of which he was ably assisted by his wife, Anne-Marie, a 
scholar in her own right, are highly regarded by the scientific community. 
The pursuit of this interest resulted in his building up one of the finest pri- 
vate scientific libraries in the country which now houses well over 5 000 vol- 
umes and 14 000 reprints covering a broad spectrum of scientific and philo- 
sophical topics. He commenced this enormously valuable collection while 
still a student when he had the leisure to attend the frequent academic book 
sales which took place in Amsterdam and Leiden, and augmented it on each 
of his return trips to Europe. The cataloguing of his reprint collection he al- 
ways carried out from 04h30-07h00 on Sunday mornings. (He invariably 
rose at 04h30 in summer and, as a concession to the longer nights, 05h00 in 
winter). 

His leadership in scientific and philosophical circles has been recognised 
by the award of a Fellowship of the Royal Society of South Africa in 1951 
and by his election to the Presidency of this august body in 1966. He also 
served the scientific community as Council member or member of numerous 
scientific bodies including the American and South African Associations for 
the Advancement of Science, Die S.A. Akademie vir Wetenskap en Kuns, 
the S.A. Association of Botanists, the British and American Mycological 
Societies and the Royal Netherlands Botanical Society. His interest in cul- 
tural affairs was very strong and he was tremendously keen on promoting 
cultural ties between Europe and South Africa. This he was able to do in 
Bloemfontein as President of the Alliance Franqaise and through his com- 
mittee work for the Algemene Nederlands Verbond; in Cape Town he was 
President of the Verbond Nederlands-Suid Afrika for a number of years. 

In spite of his scientific achievement, Wilhelm will probably be best re- 
membered by his students for the excellence of his teaching. As one of the 
country’s best-read scientists Wilhelm had an extraordinary wealth and 
depth of knowledge on which to base his lectures. This knowledge he com- 
municated to his students with a clarity and humour which, together with his 
gentleness of manner and compassion for those less gifted than himself, 
made him a deeply loved and respected lecturer and councillor. 

In the passing of Wilhelm Liitjeharms the scientific community of South 
Africa has lost a gifted scholar, researcher and philosopher who has served 


327 


it well. He is survived by his widow, Anne-Marie, his son Dr. Johann 
Lutjeharms, already a prominent scientist like his father, his daughter, Mrs. 
Marijke Haarhoff, and his grandchildren. 


O. A. M. Lewis 







J1 S. Afr. Bot. 49 (4): 329-335 ( 1983) 


TWO NEW SPECIES OF PSORALEA (FABACEAE) IN SOUTH AFRICA 
Charles Stirton 

(The Herbarium , Royal Botanic Gardens, Kew, Richmond, Surrey TW9 
3AE, England ) 

Abstract 

Two new species of Psoralea L. are described. One of these, P. implexa 
C.H. Stirton, is recorded from the Mountain Fynbos vegetation type on Du Toit’s 
Peak, near Worcester; and the second, P. trullata C.H. Stirton, from the southern 
Cape Mountains. 

UlTTREKSEL 

TWEE NUWE PSORALEA (FABACEAE) SOORTE IN SUID-AFRIKA 

Twee nuwe Psoralea- soorte word beskryf. Die een, P. implexa C.H. Stirton is in 
Bergfynbos plantegroei op Du Toitspiek naby Worcester en die tweede, P. trullata 
C.H. Stirton, is op die suidelike Kaapse berge gevind. 

Key words: Psoralea, sp. nov., Fabaceae, southern Cape, south western Cape. 
Introduction 

In 1930 Miss H. M. L. Forbes published her revision of 49 species of 
Psoralea L. in southern Africa. Since that time many new areas have been 
opened up to exploration and it is therefore not surprising that some new 
species have been discovered. 

This paper is a tribute to Elsie Esterhuysen whose collections of Psora- 
leoid legumes in South Africa are of critical importance to our understand- 
ing of Psoralea and Otholobium C.H. Stirton. The two new species described 
here are rare and unusual representatives of the genus Psoralea. 

Psoralea implexa C.H. Stirton, sp. nov.; P. trullatae C.H. Stirton affinis 
sed habitu graminoso, caulibus complanatis, floribus solitariis latericiis vel 
salmoneis, vexillolate elliptico et cupula brevite trifida differt. 

Planta perennis rhizomata acervis implexos formans. Caules complanate, 
angulati glabrescentes. Folia unifoliolata. Foliola linearioblonga vel falcata, 
apice basique acuminatis. Flores solitarii axillares 12 mm longi laterici vel 


Accepted for publication 12th April, 1983. 


329 


330 


Journal of South African Botany 



Two new species of Psoralea (Fabaceae) 


331 


salmonei. Cupula breviter trifida. Calyx accrescens. Vexillum late ellipticum. 
Carina alas superans apice rotundata. 

Type: CAPE — 3319 (Worcester): Delabat Ravine, NE ridge of Du Toit’s Peak 
(-CA), 7/2/1975, Esterhuysen 33765 (NBG, holotype; BOL, NBG, PRE, isotypes). 

Rhizomatous multi-stemmed perennial growing in tangled heaps, with 
some trailing branches up to 1,7 m long. Stems flattened, angled, glabres- 
cent. Leaves unifoliolate. Stipules 3-6 mm long, 1-3 mm wide, fused at 
base, adnate to base of petiole, falcate; blade glabrous, glandular, margins 
papillose or hairy. Petioles 1 mm long, glabrous. Leaflets 35-60 mm long, 
2-3 mm wide, linear-oblong to falcate, basal leaves shorter and up to 4 mm 
wide; glabrous, glandular; apex acuminate, base acuminate. Inflorescence 
reduced to a single flower, axillary, forming near the ends of lateral 
branches. Flowers 12 mm long, brick-red to salmon. Cupulum shortly trifid. 
Calyx 8 mm long; teeth equal, 8 mm long, 1, 8-2,0 mm wide; vexillar lobes 
fused for three-quarters their length, incurving; tube 3 mm deep; lobes gla- 
brous outside, finely ciliate on margins of the teeth and on inner face of 
teeth; accressent in fruit. Standard 12 mm long, 8 mm wide, broadly elliptic, 
weakly auriculate, appendages absent; glabrous; apex convex, slightly 
beaked; claw 3 mm long. Wing petals 10 mm long, 3 mm wide at widest 
point, 1,5 mm wide below weak auricle; claw 3, 5-4,0 mm long; sculpturing 
upper basal comprised of 8-12 irregular, reclined lamellae. Keel blades 
9, 0-9, 5 mm long, 3,5 mm wide, rounded at apex and beaked below, shorter 
than the wing petals; claw 4 mm long. Androecium 9 mm long, 2 mm wide 
at base, split; vexillar stamen lightly fused for the middle part of its length. 
Pistil 9 mm long; ovary 1,5 mm long, glabrous except for a few glandular 
hairs on the sides; gynophore present, 1 mm long; height of curvature 
2,5 mm; style glabrous, somewhat thickened before the point of flexure; 
stigma capitate, finely penicillate. Fruit 6 mm long, 4 mm wide, papery, re- 
ticulate. Seed brown black, 5 mm long, 3 mm wide (Fig. 1). 

This species is so distinctive that I have no hesitation in describing it as 
new. 

Specimens Examined 

Type only, as above 


Fig. 1. 

Psoralea implexa (Esterhuysen 33765)'. 1. Flowering branch; 2. Leaflet; 3. Undersur- 
face of leaflet; 4. Transverse section of leaflet; 5. Stipules; 6. Transverse section ot 
stem; 7. Calyx; 8. Cupulum; 9. Lateral view of seed; 10. Hilar view of seed; 11. Stan- 
dard; 12. Keel blade; 13. Wing petal. 14. Androecium. 15. Pistil. (1, 4, X 1; 2, x 2; 
2,3,7,9-15, x 3; 5, 6, x 5; 8, x 8) 


332 


Journal of South African Botany 


Distribution and Biology 

Psoralea implexa is a rare mountain endemic collected for the first and 
only time just six years ago (Fig. 2). It was found growing by Miss Elsie 
Esterhuysen at 650 m on steep open rocky stony slopes of kloofs with west- 
ern and eastern aspects. Flowering takes place in early February. Nothing is 
known about its biology. 



Known distribution of Psoralea implexa (%) and P. trullata (A) in South Africa. 
Diagnostic Features 

Grass-like habit, flattened stems, solitary brick-red to salmon-coloured 
flowers, broadly elliptic standard and shortly trifid cupulum. 

Psoralea trullata C.H.Stirton, sp. nov.; P. restioides Eckl. & Zeyhr. et 
P. implexae C.H.Stirton atque P. glaucinae Harv. affinis sed vexillo trulli- 
forme, cupulae dentibus longioribus angustioribus, carina dentibusque caly- 
cis alabastro valde attenuatis. 

Suffrutex implexa multicaulis usque 300 mm alta. Caules usque 2 mm 
longi glabri manifeste glandulosi. Folia trifoliolata inferiores unifoliolata ba- 
sales plani latique superne gradatim trifoliolata conduplicata angustioribus, 
foliola linearia vel linearioblonga basi cuneata apice acuminata. Flores soli- 
tarii axillares 10-12 mm longi purpureo-coeruli. Cupula 3-5 mm longa, tubo 
1,5 mm longo, trifida, duobus dentibus interioribus 4, 0-4, 5 mm longis an- 
guste triangularibus, dente exteriori rotundato quam interiores 4-5-plo bre- 


Two new species of Psoralea (Fabaceae) 


333 


viori. Calyx accrescens. Vexillum latissime triangulare vel trullatum. Carina 
manifeste attenuata alas prominente auriculatas valde superans. 

Type: CAPE— without precise locality: Blaauboschbaai, Fourcade 2827 (PRE, 
holotype; STE, isotype). 

Tangled multi-stemmed suffrutex, up to 300 mm tall. Stems as much as 

2 m long, glabrous, ribbed, prominently glandular. Leaves trifoliolate, oc- 
casionally unifoliolate below. Stipules 5-7(10) mm long, adnate to the pe- 
tiole, tightly clasping the stem, fused for three-quarters their length; teeth 
narrowly triangular, 2, 0-4, 5 mm long. Petioles 5-12 mm long. Leaflets 
10-16 mm long, 0, 5-2,0 mm wide; basal leaves flat and broader becoming 
conduplicate and narrower towards the apical regions; glabrous, linear to 
linear-oblong; base cuneate, apex acuminate; laterals equal to or somewhat 
shorter than the terminal leaflet. Inflorescence reduced to a single flower, ax- 
illary, occurring in the last 2-5 axils. Flowers 10-12 mm long, purplish-blue, 
subtended by two unequal bracts 2-3 mm long; pedicel 2-5 mm long. Pe- 
duncle 10-25 mm long; cupulum 3-5 mm long with 1,5 mm deep tube, 
trifid, inner pair of teeth 4, 0-4, 5 mm long, narrowly triangular, outer tooth 
rounded, 4-5 times shorter than the inner pair, 1,0-1, 5 mm long. Calyx 
12 mm long; upper four teeth equal, 8 mm long, keel tooth broader and 4 mm 
long, upcurving, vexillar lobes fused for two-thirds their length; tube 2,5- 
3,0 mm deep, ribbed; lobes glabrous inside and outside, ciliate along margins of 
the teeth, densely glandular, accressent in fruit. Standard 13-14 mm long, 
8 mm wide, very broadly triangular to trullate; claw 2,5 mm long; scarcely 
auriculate, appendages obscure, may be absent. Wing petals 14-15 mm long, 

3 mm wide, claw 3 mm long; prominently auricled, thumb-like, 1,5 mm 
long. Keel blades 20-21 mm long, slender, narrowly beaked, exceeding the 
wing petals; claw 4 mm long. Androecium 6 mm long, vexillar stamen free. 
Pistil 6 mm long; ovary 2 mm long; gynophore present, 0,5 m long; height of 
curvature 1,5— 1,6 mm long; style glabrous, thickened at point of flexure; stig- 
ma capitate, penicillate. Fruit 5-6 mm long, 3 mm wide, black, paper-thin, 
broadly reticulate. Seeds 4 mm long, 2,5 mm wide, blackish-brown (Fig 3). 

In the past this species has been referred to either P. glaucina Harv. or 
to P. restioides Eckl. & Zeyhr. It is easily separated from those species how- 
ever by its prominent elongated buds, the trullate standard, and very distinc- 
tive keel. 


Specimens Examined 

CAPE— 3322 (Oudtshoorn): Tolberg, top of Outeniqua Pass (-CD), 8/12/1962, 
H. C. Taylor 4469 (STE). 


334 


Journal of South African Botany 



2 


Two new species of Psoralea ( Fabaceae ) 


335 


— 3323 (Willowmore): de Hoek farm, Langkloofberge, 22/2/1945, Fourcade 6535 
(BOLL Helpmekaar Peak (-DC), 28/1/1941, Esterhuysen 4604 (BOL); Formosa 
Peak, Tsitsikamma Mountains, 30/1/1941, Esterhuysen 4651 (BOL); Cradock Peak, 
Montagu Pass, 1/1940, Stokoe s.n. (SAM 54873); Camel Pile, Tsitsikamma Moun- 
tains (-CD), 12/1/1947, Esterhuysen 13573 (BOL); Tsitsikamma Mountains near Jou- 
bertina, 26/1/1950, Esterhuysen 16764 (BOL, PRE). 

— 3324 (Steytlerville): Witte Els Bosch Peak (-CD), 16/11/1941, Esterhuysen 6789 
(BOL); Cockscomb Peak, Great Winterhoek Mountains (-DB), 3/3/1957, Esterhuy- 
sen 27144 (BOL). 

Without precise locality: Blaauboschbaai, Fourcade 2827 (PRE, STE). 

Distribution and Biology 

Psoralea trullata is endemic to the southern Cape and is found growing 
on the Tsitsikamma, Langkloof, Outeniqua and Great Winterhoek Moun- 
tains (Fig. 2), between the altitudes of 1 100 and 1 600 m. Very little is 
known about the habitat preferred by this species except that it is associated 
with steep rocky slopes and ledges (Esterhuysen 27144), apparently on peaty 
soils overlying TMS rock (Taylor 4469). Flowering takes place between 
November and February. 

Diagnostic Features 

Tangled herb; trullate standard; elongated flower bud; trifid unequally 
toothed cupulum; narrowly attenuated keel-blade exceeding the prominent- 
ly auricled wing petals; accrescent calyx. 

Acknowledgements 

The author wishes to acknowledge the assistance and facilities made 
available to him at the Botanical Research Institute, Pretoria and Stellen- 
bosch, the Compton Herbarium and the Bolus Herbarium. The illustrations 
were drawn by Mrs. Rosemary Holcroft of the Botanical Research Institute, 
Pretoria and are gratefully acknowledged. Miss C. M. Wilmot-Dear kindly 
provided the Latin diagnoses. 

References 

Forbes, H. M. L., 1930. The genus Psoralea Linn. Bothalia 3: 116-136. 


Fig. 3. 


Psoralea trullata (Fourcade 2827): 1. Flowering branch, 2 Calyx m i fruit, 3 C ly , 
4. Cupulum; 5. Bracts subtending each flower in a cupulum 6 Bud 7. Standam 

8 Keel blade; 9. Wing petal; 10. Pistil; 11. Androecium. (1, X 1, 2, ’ 

x 4 ; 4, x 5; 5, x 8). 







J1 S. Afr. Bot. 49 (4): 337-342 (1983) 


TWO NEW SPECIES OF OTHOLOBWM (FABACEAE) 

Charles Stirton 

(The Herbarium , Royal Botanic Gardens, Kew, Richmond, Surrey TW9 
3AE, England ) 

Abstract 

Two new species of Otholobium Stirton are described. One of these, O. rubicun- 
dum C.H. Stirton, is recorded from the transition belt of Succulent Mountain Scrub 
and False Macchia on the Klein Swartberg Mountains; and the second, O. pictum 
C.H. Stirton, from the Baviaanskloof Mountains. 

UlTTREKSEL 

TWEE NUWE OTHOLOBIUM (FABACEAE) SOORTE 

Twee nuwe Otholobium Stirton-soorte word beskryf. Een van die soorte, O. ru- 
bicundum C.H. Stirton, word gevind in die oorgang strook van Sukkulente Berg- 
struikveld en Skyn-Fynbos op die Klein Swartberge en die tweede, O. pictum 
C.H. Stirton op die Baviaanskloofberge. 

Key words: Otholobium, sp. nov., Fabaceae, southern Cape, eastern Cape. 

Otholobium rubicundum C.H. Stirton, sp. nov. 

Frutex parvus, effusus. Foliola trifoliolata. Foliola 15-25 mm longa, 4-6 
mm lata, conduplicata, anguste ovovata, glabrescentia, valde recurvato-mu- 
cronata, basi cuneata; lateralia breviora. Inflorescentia 3-6 flora, axillaris. 
Flores 9-10 mm longi, rosei. Dentes calycis aequales. Vexillum late ovatum, 
vix auriculatum, unguiculatum. Alae carina longiores. 

Type: CAPE — 3321 (Ladismith): Hoeko road, south base of Klein Swartberg 
(-AD), 10/1957, Barker 1606 (NBG, holotype). 

Small spreading shrublet. Stems slender, ridged, waxy, sparingly covered 
with glands and small appressed white hairs. Flowering shoots axillary, ap- 
pressed pubescent with numerous elliptic raised pustules. Leaves trifoliolate. 
Stipules 2, 5-3, 5 mm long, 1, 5-2,0 mm wide, triangular; papery, sparingly 
pubescent but covered with orange-coloured, sunken, rounded glands. Pe- 
tioles < 1 mm long, hairy. Leaflets 15-25 mm long, 4-6 mm wide, condupli- 
cate, glabrescent, glandular, narrowly obovate, base cuneate, apex strongly 


Accepted for publication 12th April, 1983. 


337 


338 


Journal of South African Botany 


recurved mucronate; laterals shorter than terminal leaflet. Inflorescence axil- 
lary, condensed, comprising 1-2 sets of triplet flowers; each set subtended 
by a single 3 mm long, 1,5 mm wide, narrowly obovate, acute, hairy and 
glandular bract. Flowers 9-10 mm long, pink; pedicel 2 mm long. Peduncle 
1-2 mm long. Calyx 1 mm long; teeth equal, margins densely ciliate, 4 mm 
long, upper four teeth 1,7 mm wide, keel tooth 3 mm wide, vexillar teeth 
slightly connate; lobes prominently glandular, glabrous inside and outside 
except for a few hairs along the ribs; tube 3 mm deep, ribbed. Standard 12 
mm long, 8 mm wide, broadly ovate, scarcely auriculate, appendages ab- 
sent; claw 4 mm long. Wing petals 10-11 mm long, 3 mm wide, auricle 1 mm 
high; sculpturing upper basal and left central comprised of 35-40 parallel la- 
mellae. Keel blades 7 mm long, 3 mm wide, blade bulging; claw 3,5 mm 
long. Androecium 7 mm long; vexillar stamen fused to sheath for two-thirds 
its length. Pistil 5,5 mm long; ovary 1,5 mm long, glandular with few hairs, 
gynophore absent; height of curvature 1 ,5 mm long; style glabrous 0,5-0, 6 mm 
thick at point of flexure; stigma minutely capitate, finely penicillate. Fruit 
and seed unknown. 

Otholobium rubicundum is very rare, having been collected for the first 
time, as recently as 1957, by Miss W. F. Barker. This species has not been 
found since, despite a thorough search by the author in 1980. 

Specimen Examined 

Same as for type. 


Distribution and Biology 

Otholobium rubicundum was found by Miss Barker at an altitude of 
about 650 m, apparently in the transition belt between False Macchia and 
Succulent Mountain Scrub vegetation types. Apart from the knowledge that 
flowering takes place during late October and early November little else is 
known about the biology of this rare and distinctive shrublet. 

Diagnostic Features 

Small spreading shrublet; leaflets conduplicate, glandular; inflorescence 
3-6-flowered, condensed; flowers pink, each triplet subtended by a narrowly 
obovate acute bract; standard broadly ovate. 


Fig. 1. 

Otholobium rubicundum ( Barker 1606): 1. Flowering shoots; 2. Stipule; 3. Flower 
bracts; 4. Calyx opened out; 5. Leaf; 6. Standard; 7. Wing petal; 8. Base of keel; 
9. Keel; 10. Pistil; 11. Androecium. (1, X 1; 2-11, x 4.) 


Two new species of Otholobium (Fabaceae) 


339 




340 


Journal of South African Botany 


Otholobium pictum C.H.Stirton, sp. nov. 

Frutex erectus usque 2 m altus. Rami luteo-fusci. Folia trifoliolata. Folio- 
la 15-25 mm longa 2, 5-3,0 mm lata plana, anguste obovata atque aliquan- 
tum falcata, prominente nigropunctata, recurvata, mucronata, basi cuneata, 
lateralia minoria. Inflorescentia 60-90-flora, dense spicata, late ovata vel ob- 
longa. Flores 9-10 mm longa, albi. Calycis dentes inaequales lobus carinae 
ceteris lobis 2 mm longior. Vex ilium late ellipticum, vix auriculatum, ungui- 
culatum. Alae carina longiores. 

Type: CAPE — 3324 (Steytlerville): Baviaanskloof mountains, between Smits- 
kraal and Wilgehof (-CB), 17/9/1973, Oliver 4588 (STE, holotype; PRE, isotype). 

Erect shrubs up to 2 m tall. Stems slender, glabrous, weakly ribbed, yel- 
lowish-brown. Flowering shoots clustered at the ends of the previous years 
shoots, densely appressed white pubescent between the angles, with elliptic 
pustules scattered along the upper edges. Leaves trifoliolate. Stipules 
2, 5-3,0 mm long, 0,5-0, 8 mm wide, fused to the base of the petiole, shortly 
triangular, scarcely pubescent, densely encrusted with orange glands espec- 
ially at the base. Petioles 1, 5-2,0 mm long, gland-encrusted. Leaflets 15-25 mm 
long, 2, 5-3,0 mm wide, flat, glabrescent, prominently nigro-punctate in 
dried specimens, narrowly obovate, somewhat falcate, base cuneate, apex 
recurved mucronate; laterals shorter and narrower than terminal leaflet. 
Inflorescence densely spicate, broadly-ovate becoming oblong, comprising 
from 20-30 sets of triplet flowers; each subtended by a single broadly 
ovate, pubescent glandular bract. Flowers 9-10 mm long, white; subsessile. 
Peduncle absent. Calyx 8 mm long; upper four teeth more or less equal, 
6 mm long, 1 mm wide, keel tooth 8 mm long, 1 mm wide; vexillar teeth par- 
tially connate; lobes covered in small glands, densely black tomentose out- 
side; tube 3, 0-3, 5 mm deep. Standard 9, 0-9, 5 mm long, 6,5 mm wide, 
broadly elliptic, emarginate, slightly auriculate, appendages absent; claw 2 mm 
long. Wing petals 9 mm long, 2, 0-2, 5 mm wide, longer than the keel, 
auriculate, claw 3, 0-3, 3 mm long; sculpturing upper basal, upper central 
and upper left distal comprised of 20-26 irregularly parallel lamellae. Keel 
blades 7 mm long, 2,5 mm wide, apex rounded; claw 3,5 mm long, purple 
blotch present on inner face of tip. Androecium 7 mm long; vexillar stamen 
6, 0-6, 5 mm long, loosely fused to sheath for half its length. Pistil 6, 5-7,0 mm 
long; ovary 2,5 mm long, pubescent; gynophore 0,5 mm long; style glab- 
rous, 0,4 mm thick at point of flexure; stigma papillose, forward-sloping. 
Fruits and seed unknown. 

Otholobium pictum was first collected in 1930. Since then it has been col- 
lected only twice. But considering the inaccessibility of the terrain in which 


Two new species of Otholobium ( Fabaceae ) 


341 


Fig. 2. 

Otholobium pictum (Oliver 4588): 1. Flowering shoot; 2. Calyx with petals removed; 
3. Calyx flattened out; 4. Leaf, surface view; 5. Tip of leaf; 6. Standard; 7. Stipule; 
8. Keel; 9. Wing petal; 10. Androecium; 11. Pistil. (1, X |; 4, X ca.l; 3, X 3; 6,8-11 

X 4; 7, x 8) 



this distinctive species occurs it is not surprising that the only collections 
made so far have been along the only two passable roads across the Ba- 
viaanskloof and Great Winterhoek Mountains. It is probably fairly common 
on the escarpment and with further collecting by mountaineers it should 
have its range broadened substantially. The discovery of this species stresses 
once again the poor state of collecting along the inland mountains east of 
Uitenhage and north of Stormsriver. 


342 Journal of South African Botany 



Known distribution of Otholobium rubicundum (A) and O. pictum (#). 
Specimens Examined 

CAPE — 3324 (Steytlerville): Baviaanskloof Mountains, between Smitskraal and 
Wilgehof (-CB), 17/9/1973, Oliver 4588 (PRE, STE); Wintershoek Mountains (-DB), 
15/9/1930, Fries, Norlindh & Weirmarck 1065 (PRE, STE); Mountain road north of 
Patensie, 12/7/1954, Taylor 1285 (PRE, SAAS). 

Distribution and Biology 

Otholobium pictum, a distinctive endemic of the Baviaanskloof Moun- 
tains, remains little known and seldom collected. It is quite surprising that 
such a distinctive shrub should have been without a name for so long. 

This species flowers during August and September. Nothing more is 
known about its biology. 

Diagnostic Features 

Erect 2 m tall shrub; leaflets flat, nigro-punctate; inflorescence 60-90- 
flowered, densely spicate; flowers white, each triplet subtended by a broadly 
ovate bract; standard broadly elliptic. 

Acknowledgements 

I would like to thank Mr. E. G. H. Oliver (Stellenbosch) for his com- 
ments about O. pictum; Miss C. M. Wilmot-Dear for the Latin diagnoses; 
Mrs. R. Holcroft for kindly drawing the plates and Dr. B. de Winter, Direc- 
tor, Botanical Research Institute, for permission to publish them. Finally I 
would like to thank the Curators of the Compton, Kew, Saasveld, Stellen- 
bosch (Government) and National Herbaria for the loan of their material. 


J1 S. Afr. Bot. 49 (4): 343-346 (1983) 


BOWIEA GARIEPENSIS: A NEW BOWIEA SPIECIES (LILIACEAE) 

FROM THE NORTH WESTERN CAPE 

E. J. van Jaarsveld 

(National Botanic Gardens of South Africa, Kirstenbosch, Private Bag X7, 
Claremont 7735, R.S.A.) 

Abstract 

A new Bowiea species (Liliaceae) is described from the north western Cape 
(R.S.A.) 

UlTTREKSEL 

BOWIEA GARIEPENSIS: 'N NUWE BOWIEA - SOORT VANAF NOORDWES- 
KAAP 

’n Nuwe Bowiea - soort (Liliaceae) vanaf Noordwes-Kaap (R.S.A.) word beskryf. 
Key words: Bowiea, sp. nov., Liliaceae, north western Cape. 

Bowiea gariepensis E.J. van Jaarsveld, sp. nov.; differt ab Bowiea volubi- 
lis Harv. quia habet prostratos ad pendulos et glaucos caules, qui ferunt con- 
spicuos et albos flores 25 mm diametris, stigmatibus 4,5 mm longis. 

Type: Cape Province — 2919 (Pofadder): Upper south slope of Groot Pellaberg 
(-AA), van Jaarsveld 6650 (NBG, holotype; PRE, isotype). 

Geophyte, glabrous in all parts. Roots fleshy, white, up to 5 mm in di- 
ameter. Bulb globose, depressed, 60-140 mm in diameter, subterranean or 
somewhat exposed, the exposed part green, often forming 2-3 bulblets; tu- 
nics thick (up to 8 mm) and fleshy; outer tunics withered, grey and membra- 
nous, completely covering the bulb. Leaves appearing before the inflores- 
cence, lanceolate, canaliculate up to 25 mm long and 10 mm wide, striate, 
withering shortly prior to flowering. Stem (inflorescence) annual, pendulous 
and scrambling, much-branched, basally flexuose, up to 1,2 m long, softly 
succulent, glaucous; branchlets subulate, 20-70 mm long, 2,5-5 mm in di- 
ameter; bracts lanceolate, spurred, 3-5 mm long, about 2 mm wide at the 
base; pedicels arched, 25-70 mm long. Flowers opening diurnally, relatively 
few, scented. Perianth 25 mm in diameter, divided to the base, segments 


Accepted for publication 20th May, 1983. 


343 


344 


Journal of South African Botany 


patent, oblong to lanceolate, 12 mm long, 3,5 mm broad with a medium 
green stripe, margins revolute towards the base; apex subacute, each peri- 
anth lasting approximately 8-10 days. Stamens suberect, about 5 mm long. 
Ovary broadly conical, 5 mm in diameter at the base, light green. Style 
terete with a 3-lobed stigma. 

Flowering period: March to June. 

Distribution and Habitat 

Bowiea gariepensis occurs in Namaqualand Broken Veld (Acocks Veld 
type no. 33) in the lower Orange River valley in Namibia and South Africa, 
from Helskloof (Richtersveld) in the west to Pellaberg in the east. 

The plants are procumbent and pendulous among rocks and occur main- 
ly on south-facing slopes, being in shade for most of the day. The stems tend 
to sprawl, often cascading over the rocks. The plants are locally abundant. 
Rainfall is mainly in autumn and winter and ranges from 50-200 mm per an- 
num. 

Discussion 

Bowiea gariepensis is a distinct species of the winter rainfall region of the 
north western Cape. The glaucous pendulous stems with the conspicuous 
white, larger corolla, and the spreading segments, clearly separate it from 
the only other member of the genus in South Africa, Bowiea volubilis Harv. 
The active growing season of Bowiea gariepensis is from April to Septem- 
ber, and the plant becomes deciduous towards the summer. In contrast, 
Bowiea volubilis , which occurs in the central and eastern parts of South 
Africa, grows during the summer months. The latter species has bright 
green branches and an inconspicuous green corolla with distinctly reflexed 
segments. In addition, the bulb of Bowiea gariepensis is covered with grey 
membranous sheaths which are lacking in Bowiea volubilis. 

Specimens Examined 

CAPE — 2817 (Vioolsdrif): Northern base of Little Helskloof ( — CB), Wisura 1462 
(NBG 1035170). 

— 2819 (Ariamsvlei): Pelladrift (-CC), Dinter 5795 (SAM); Warmbad (-BD), Range 
696 (SAM). 

Acknowledgements 

Miss Deirdre Snijman is thanked for help and for editing the manuscript; 
Mrs. Ellaphie Ward-Hilhorst for her illustration, and Mr. Norval Gelden- 
huys for the preparation of the Latin diagnosis. 



Bowiea garipiensis: a new Bowiea species 


Fig. 1. 

Bowiea gariepensis van Jaarsveld, sp. nov. 


346 


Journal of South African Botany 


References 

Baker, J. G., 1896. In: Thistleton-Dyer, W. T. (ed.), Flora Capensis 6: 367. 

Dyer, R. A., 1976. The Genera of Southern African Flowering Plants 2: 927. Pre- 
toria: Government Printer. 


J1 S. Afr. Bot. 49 (4): 347-422 (1983) 


THE AFRICAN GENUS TRITON I A KER-GAWLER: PART 2. SECTIONS 
SUBCALLOSAE AND MONTBRETIA 


Miriam P. de Vos 

(Department of Botany, University of Stellenbosch , Stellenbosch 7600, 
R.S.A.) 

Abstract 

The remaining two sections of the genus Tritonia, namely Subcallosae sect, nov., 
with five species, and Montbretia (DC.) Pax with 14 species, are dealt with, conclud- 
ing the revision of the genus, with a total of 28 accepted species. These two sections 
are widespread, mainly in the summer rainfall regions of South Africa, and also in 
several African countries south of the equator. 

Illustrated descriptions of the species, their range and synonymy, as well as keys, 
are provided. In section Subcallosae two new varieties are described and three 
species are reduced to subspecific rank. Four new species are described in section 
Montbretia, namely T. delpierrei, T. drakensbergensis, T. karooica and T. marlothii. 
A list is given of ca. 50 species excluded from the genus, and a few insufficiently 
known species are recorded. There is also a note with newly acquired data on T. tug- 
welliae that was treated in Part 1. 

UlTTREKSEL 

Die oorblywende twee seksies van die genus Tritonia, naamlik Subcallosae sect, 
nov., met vyf spesies, en Montbretia (DC.) Pax met 14 spesies, word behandel, waar- 
deur die hersiening van die genus afgesluit word, met ’n totaal van 28 aanvaarde spe- 
sies. Die twee seksies kom wydverspreid voor, hoofsaaklik in die somerreenvalstreke 
van Suid-Afrika, en ook in sekere Afrikalande suid van die ewenaar. 

Gei'llustreerde beskrywings van die spesies, hul sinonimie en verspreiding, word 
gegee en sleutels word verskaf. In seksie Subcallosae word twee nuwe varieteite be- 
skryf en drie spesies word tot subspesifiese rang verlaag. Vier nuwe spesies word vir 
seksie Montbretia beskryf, nl. T. delpierrei, T. drakensbergensis, T. karooica en 
T. marlothii. ’n Lys word gegee van die ongeveer 50 uitgesluite spesies, en enkele 
onvolledig bekende spesies is opgeteken. Daar is ook ’n nota met nuwe feite oor T. 
tugwelliae wat reeds in Deel 1 behandel is. 

Key words: Tritonia, sect. Subcallosae, sect. Montbretia Iridaceae, South Africa. 

Introduction 

In this article the remaining two sections of the genus Tritonia, namely 
sections Subcallosae and Montbretia, are revised. The species of these two 


Accepted for publication 30th May, 1983. 


347 


348 


Journal of South African Botany 


sections are widely distributed in the summer rainfall regions of South Africa, 
as well as in regions which, although often incorporated with the winter 
rainfall region, may also receive rain in any season of the year (southern 
Cape Province, Little Karoo and Namaqualand (Figs 15, 16, 21, 22). Unlike 
sections Tritonia and Pectinatae , they do not extend into the winter rainfall 
region proper, i.e., the extreme south western Cape Province. Only two 
species, namely T. moggii and T. laxifolia, occur beyond the borders of 
South Africa, as far north as Zambia and Tanzania. No species of Tritonia 
occurs in West Africa, and the few species north of the equator which were 
described as Tritonia , belong to other genera. 

In both sections now under consideration, the length of the perianth tube 
varies from species to species, being shorter than the segments in a few 
species, equal to in others, and up to several times the length of the seg- 
ments in still others. The degree of zygomorphy of the flowers differs in the 
two sections, as well as the development, size and shape of the outgrowths, 
known as calli, on the anticous perianth segments. 

Section Subcallosae , established here, incorporates species which some- 
times show small vestigial calli, usually in the form of low ridges, on one or 
sometimes all three anticous perianth segments. Its species have flowers with 
slight zygomorphy. The section can be regarded as the most primitive of the 
genus, forming a link with the other sections. 

Section Montbretia comprises the most highly developed species of the 
genus. The flowers show a decided zygomorphy, as in sections Pectinatae 
and Teretifolia, and well-developed calli, 1-5 mm high, occur on the three 
anticous perianth segments, or rarely on the median one only. 

Montbretia was established as a genus by A. P. De Candolle (1803) for 
two species of Gladiolus with calli. There was uncertainty about the rank of 
this taxon from the time that Bentham and Hooker filius (1883) brought it to 
the synonymy of Tritonia. See further under section Montbretia. 

A revised key to the five sections is again given, to help in distinguishing 
the sections. 

Key to the Sections 

1. Flowers almost regular (rarely completely regular - T. dubia) with spreading 
perianth segments; stamens randomly curved (rarely erect), with anthers not 
contiguous and not all facing the anterior side of the flower: 

2. Perianth tube shorter than or sometimes subequal to the segments; leaves 
lanceolate, ensiform, flat 1. Sect. Tritonia 

2. Perianth tube longer than the segments; leaves linear, striate (when dry) 

and subterete or compressed cylindrical (fresh) 4. Sect. Subcallosae 

1. Flowers zygomorphic with the posticous perianth segment usually widest; sta- 
mens curved towards the posticous segment with anthers contiguous, facing 
the anterior side of the flower: 

3. Spike dense, curved slightly at its base to give it an almost pectinate ap- 


African genus Tritonia Ker-Gawler: part 2 


349 


pearance; flowers white, cream or pale pink with reddish honey guides in 
the throat; leaves with crisped-undulate or inrolled margins or with four 

longitudinal ribs or flanges “ 2. Sect. Pectinatae 

3 . Spike lax or sometimes dense, erect or suberect, secund or distichous, not 
pectinate: flowers red to orange, yellow or sometimes pale, often with yel- 
low in the cup, sometimes with a callus on the lower perianth segments; 
leaves linear to lanceolate or ensiform, rarely undulate or filiform: 

4. Leaves filiform, terete or subterete, wiry, striate when dry 

3. Sect. Teretifolia 

4 . Leaves linear to lanceolate or ensiform, rarely spirally twisted with un- 
dulate margins: 

5. Perianth with a well-developed callus on the three, or sometimes 

only the median, anterior segments 5. Sect. Montbretia 

5. Perianth without calli or with a low ridge on the median or rarely all 
three anterior segments 4 Sect. Subcallosae 

Key to Section Subcallosal 

1. Leaf margins straight: 

2. Leaves compressed cylindrical to subterete, slightly swollen (fresh), linear 
and striate (dry); flowers almost regular with long perianth tube; stamens 

almost central with anthers facing outwards 10. T. bakeri 

2. Leaves linear-lanceolate to linear, thin and flat, with a prominent middle 
vein and sometimes a prominent vein near each margin; flowers zygomor- 
phic with long or short perianth tube; stamens curved towards the posterior 
side of the flower, with anthers unilateral, facing the anterior side: 

3. Perianth tube at least 1,5 times the length of the segments; leaves with- 
out a prominent vein near the margins: 

4. Outer bract, of the lower flowers especially, acuminate or rarely 
with three acuminate tips close together 12. T. flabellifolia 

4 . Outer bract 3- or sometimes 2-toothed or irregularly toothed 

11. T. pallida 

3 . Perianth tube subequal to, or slightly longer or shorter than the seg- 
ments; leaves with a prominent vein near each margin: 

5. Flowers cream, straw-coloured or pale yellow, usually with dark 

veining, sometimes with apricot or orangy tints towards the perianth 
tips or outside; flowering period August to November (to Decem- 
ber) 13. T. lineata 

5. Flowers red. orange-red or pink, rarely with dark veining; flowering 

period (December to) January to April 14. T. disticha 

1 . Leaf margins undulate T. undulata 

(see under Species Insufficiently Known) 


Key to Section Montbretia 

1. Leaves with undulate margins, sometimes spirally twisted ... 20. T. watermeyeri 
1 . Leaves with straight margins, not spirally twisted: 

2. Perianth tube (l,5-)2-3(-4) times the length of the segments: 

3. Flowers somewhat salver-shaped at noon, with almost equal perianth 
segments; leaves slightly succulent; Richtersveld, Namaqualand. 

4. ''Style 18-25 mm long; perianth tube 1,5-2 times the length of the _ 
segments ^5. T. e pierrei 

4 . Style 32-40 mm long; perianth tube 3-4 times the length of the seg- 

26- T. marlothu 


350 


Journal of South African Botany 


3. Flowers somewhat 2-lipped or funnel-shaped, with the posticous peri- 
anth segment often widest; leaves not succulent; southern and eastern 
Cape Province, Karoo and Mozambique: 

5. Perianth red, orange-red or pink: 

6. Perianth tube less than 20 mm long, 1,5 to almost 2 times the 
length of the segments; perianth dark red, drying to a blackish 
purple-red; eastern Cape Province 23. T. atrorubens 

6. Perianth tube more than 20 mm long, about twice the length of 

the segments; perianth salmon-pink or orange-pink; Mozam- 
bique 24. T. moggii 

5. Perianth bright yellow or brownish-yellow, sometimes apricot or 
“cream and orange”: 

7. Leaves suberect or sometimes spreading; perianth bright yellow, 

somewhat 2-lipped; posticous segment erect (porrect) and other 
segments spreading; southern and south eastern Cape 

16. T. chrysantha 

7. Leaves usually strongly reflexed; perianth brownish-yellow, apri- 
cot or “cream and yellow”, with dark veins, funnel-shaped, seg- 
ments often reflexed; Great Karoo 27. T. karooica 

2. Perianth tube nearly equal to or (slightly) shorter than the segments: 

8. Scape not extending above leaf bases; inflorescence with 2-4 flowers 
each terminal on a short stalk hidden by leaf bases; leaves reflexed 

28. T. florentiae 

8. Scape extended, usually longer than leaves; inflorescence a simple or 
branched spike with several flowers; leaves suberect or spreading: 

9. Leaves linear, (0,5-)2-4(-6) mm wide: 

10. Plants not more than 300 mm long; leaves somewhat flaccid; 

southern Cape 17. T. parvula 

10. Plants (250-)300 — 900 mm long; leaves firm; northern Cape and 
Transvaal: 

11. Perianth tube 8-10 mm long, about half the length of the 
segments: leaves with a prominent vein near each margin; 

north eastern Cape 21. T. drakensbergensis 

11. Perianth tube ( 10— )14— 18 mm long, subequal to or slightly 
longer or rarely shorter than the segments; leaves striate 

when dry; Transvaal 22. T. nelsonii 

9. Leaves lanceolate, ensiform or linear-lanceolate, with widest leaves 
usually more than 5 mm in width (except T. strictifolia ): 

12. Flowering period spring or early summer; capsules 7-10 mm 
long: 

13. Leaves usually erect, (40-)60-90 x 2-5 mm, with a promi- 
nent vein near each margin; callus on the anticous perianth 

segments high and narrow 18. T. strictifolia 

13. Leaves suberect or spreading, longer than 100 mm, usually 
wider than 5 mm, usually without a prominent vein near the 
margins, calli disc-shaped (as high as long), rarely high and 

narrow 15. T. securigera 

12. Flowering period autumn, sometimes late summer or early win- 
ter; capsules 10-15 mm long, erect, pressed against the rachis 
19. T. laxifolia 


African genus Tritonia Ker-Gawler: part 2 


351 


4. Section SUBCALLOSAE DeVos, sect. nov. 

Folia lanceolata vel linearia, interdum subteretia, subsucculenta, inter- 
dum nervo prominenti prope margines. Flores zygomorphi vel interdum fere 
regulares, plerumque infundibulares, rubri vel miniati vel flavi vel albi, tubo 
perianthii infundibulari vel pro parte majore tubiformi, parum breviore vel 
ad 2-3-plo longiore segmentis; saepe segmento antico medio vel interdum 
omnibus tribus segmentis anticis callo vestigiali vel crista humili praeditis. 
Stylus et stamina segmentum posticum versus curvata vel interdum fere ad 
centrum posita. 

Type species: Tritonia lineata (Salisb.) Ker-Gawl. 

Leaves lanceolate to linear, sometimes subterete and slightly succulent, 
sometimes with a prominent vein near the margins. Flowers zygomorphic or 
sometimes almost regular, usually funnel-shaped, red, orange-red, pink, yel- 
low or white. Perianth tube funnel-shaped or for the greater part tubular, 
slightly shorter to 2-3 times longer than the segments; segments subequal, 
oblong to obovate-oblanceolate, usually spreading, often with a vestigial cal- 
lus or low ridge on the median anticous, or sometimes on all three anticous 
segments. Style and stamens curved towards the posticous segment or some- 
times almost central. Anthers facing the anterior side of the flower or some- 
times outwards. 

Distribution: Widely distributed in southern districts of the Cape Prov- 
ince from Caledon to George and in the Little Karoo and Langkloof; also in 
the eastern parts of South Africa from Port Elizabeth, throughout the 
Transkei and Natal to the south eastern border of Transvaal and north east- 
ern Orange Free State, as well as in Lesotho and Swaziland: not extending 
into the Great Karoo or the south western corner and western part of the 
Cape Province (Figs 15, 16). 

Three species, namely T. bakeri, T. pallida and T. flabellifolia , mainly 
from the southern Cape Province, are winter rainfall species, sprouting in 
autumn and flowering in spring and early summer. T. lineata is widely dis- 
tributed in the eastern summer rainfall regions of South Africa and also oc- 
curs within the eastern limits of the winter rainfall region, it, too, has the 
winter rainfall habit of sprouting in autumn and flowering in spring and early 
summer. The closely related T. disticha which is more or less sympatric with 
T. lineata , flowers in summer. 

The species T. undulata is known only from its type specimen in Bui 
man's collection in Geneva, with a locality indicated merely as Caput bonac 
Spei. It perhaps fits best in this section. See under Species Insufficiently 
Known. 

In three species, namely T. pallida, T. lineata and T. disticha , a sma 




352 


Journal of South African Botany 




Fig. 15 

Geographical distribution of Tritonia, section Subcallosae: T. pallida, T. flabellifolia, 

T. bakeri. 



African genus Tritonia Ker-Gawler: part 2 


353 




Fig. 16 

Geographical distribution of Tritonia, section Subcallosae: T. lineata, T. disticha. 



354 


Journal of South African Botany 


vestigial callus, usually in the form of a low ridge, is often present on the 
median, or sometimes all three, anticous perianth segments. The perianth 
tube is short in the two eastern species and long in the southern Cape 
species. 

The section seems to form a link with the other sections of the genus and 
its two eastern, widely distributed species may possibly be considered the 
most primitive species of the genus. The perianth is not as strongly zygomor- 
phic as in sections Montbretia , Pectinatae or Teretifolia. 

The section comprises six out of the 20 species of Baker’s subgenus Tri- 
tonia Proper (1896). Ten of this score of species, however, do not belong to 
Tritonia , and the remaining four species in the subgenus have now been re- 
moved to three other sections of Tritonia. 

Leaf anatomy. Some interesting anatomical features are shown within 
this section. In T. bakeri the long, narrow leaves are compressed cylindrical 
or subterete, storing a small amount of water. The epidermis of the leaf 
margins is not palisade-like and thickened as in most species of Tritonia (De 
Vos, 1982a). In T. lineata and T. disticha the leaves have, in addition to 
their prominent middle vein, a strong vein near each margin, and the epi- 
dermis contains large amounts of tanniniferous material often causing the 
leaves to become dark brown on drying out. The strong submarginal veins 
remain pale and are especially prominent in the dry state. 

10. Tritonia bakeri Klatt in Abh. naturf. Ges. Halle 15: 358 (1882), non 
Klatt 1895. 

Plants (270-)400-700(-800) mm long. Corm depressed globose, 15- 
30 mm diam., tunic fibres fine, irregularly reticulate, subparallel in the lower 
part. Scape simple or branched, suberect or slightly flexuose, 
(2Q0-)400-700(-800) mm long, with a collar of old leaf bases around its 
base. Leaves 4-7 or more, compressed cylindrical or subterete, water stor- 
ing, rigid, acute or acuminate, erect or suberect, (150-)200-300(-600) x 
1— 2(— 3) mm, striate when dry, sometimes as long as the scape; cauline 
leaves 1-2, subulate, to 150 mm long. Spike l-2(-3)-branched, distichous, 
lax, with (3— )6— 10 flowers. Bracts membranous, papery, with brown and 
sphacelate upper part, (6— )10— 15(— 18) mm long; outer 2-3-toothed, or ir- 
regularly toothed or obtuse, the median tooth sometimes reduced, veined; 
inner bidentate or bifid with acuminate teeth, 2-veined, with papery mar- 
gins. Flowers almost regular, often somewhat salver-shaped, suberect, 

Fig. 17 

Tritonia bakeri subsp. bakeri : a, plant and front and side view of flowers; b, outer 
(left and lower) and inner (right) bract; c, flower half in a median cut. ( De Vos 2413, 

2414). 


African genus Tritonia Ker-Gawler: part 2 


355 



356 


Journal of South African Botany 


(35-)50-75(-90) mm long, cream, creamy-yellow, or mauvish-pink, the 
three outer segments with dark veins outside. Perianth tube (20-)30-45(-55) 
mm long, tubular but widening in the upper quarter to 7 mm diam., almost 
straight; segments oblanceolate-oblong or oblanceolate-obovate, obtuse to 
subobtuse, spreading, subequal in length, (15-)20-25(-30) x (5-)7-10 mm, 
with the posticous segment 8-14 mm wide. Stamens erect, almost in the 
centre of the flower; filaments (8-)15-20(-25) mm long, erect; anthers 5-10 
mm long, straight or curved, facing outwards, violet, reaching halfway or al- 
most to the tips of the perianth segments; pollen violet or pale. Style 
(27-)35-60(-80) mm, erect; stigmatic branches 3-5 mm long, reaching or 
overtopping the anthers. Capsules obovoid, 12 mm long; seeds black, 2 mm 
diam. Chromosome number 2n = 22. 

Flowering period: October to early December. 

Distribution: chiefly in the Little Karoo and Langkloof, also in the Cape 
south coastal districts from George to Riversdale (Fig. 15). 

This species is distinguished by its narrow compressed-cylindrical or sub- 
terete leaves which store some water and become striate when dried out, 
and by its lax, distichous spike with almost regular, long-tubed, pale flowers. 
Morphologically it stands nearest T. pallida from which it differs in its nar- 
rower leaves, nearly regular flowers with the stamens and style placed al- 
most symmetrically in the centre of the flower, and in its perianth without 
calli. Its distribution overlaps that of T. pallida in the Ladismith district. 
From T. cooperi, which also has narrow leaves, it differs in its lax distichous 
spike with its almost regular flowers and in its leaves which are not X-shaped 
or H-shaped in transverse section. 

The younger Linnaeus’s brief description of Gladiolus longiflorus (1781), 
based on a Thunberg specimen, seems to match T. bakeri. There are no 
Thunberg specimens so named in the Linnaean herbarium (LINN), but only 
specimens of Sparmann which are, however, Acidanthera (N. E. Brown, 
1928). One of Thunberg’s two sheets in UPS, named G. longiflorus a and (3 
respectively (nos 1040 and 1041 ) must therefore be regarded as the type. 
They are, however, two species, a being T. bakeri and (3 T. cooperi. Lin- 
naeus’s diagnosis is so short that it fits both a and (3 equally well. Three 
years later Thunberg gave a more detailed description of G. longiflorus in 
Dissertatio de Gladiolo (1784), which fits his specimen no. 1041, i.e. 
T. cooperi, best. This then was chosen as lectotype of G. longiflorus (De 
Vos, 1982b). 

Specimens collected by Oldenburg and Masson at the Cape were named 
G. striatus by Solander, a manuscript name which Baker (1876) published as 
Montbretia striata. When Klatt (1882) transferred this species to Tritonia, he 
could not use the combination T. striata for it, as this was a later homonym 


African genus Tritonia Ker-Gawler: part 2 357 

of T. striata (Jacq.) Ker-Gawler (1805) (which is Babiana striata). He there- 
fore renamed the species T. bakeri. But in 1895 he used the name T. striata 
for it, and identified the central African species which Baker (1875) had de- 
scribed as Montbretia laxifolia Klatt, as T. bakeri. Klatt’s T. bakeri of 1882 
and 1895 are therefore two different species. 

The type of T. bakeri at BM (sub Montbretia striata) comprises three 
plants with flowers which are rather small compared with numerous later 
collections of the species, and which resemble the type of T. lilacina to some 
extent. The latter differs in its pale mauvish-pink perianth and in the slightly 
smaller size of the floral organs. In some herbaria specimens of T. bakeri 
were identified as T. lilacina. The two species are so closely related that they 
are now regarded as conspecific, and T. lilacina is transfered to T. bakeri as 
subspecies lilacina. 


Key to the Subspecies 

Flowers (50— )65— 80(— 90) mm long, cream or creamy-yellow a. subsp. bakeri 

Flowers 35-50 mm long, pale mauvish-pink or sometimes cream . . . b. subsp. lilacina 


a. subspecies bakeri 

Tritonia bakeri Klatt in Abh. naturf. Ges. Halle 15: 358 (1882), non 
Klatt 1895; Bak. Handb. Irid. 193 (1892) & in FI. Cap. 6: 123 (1896) in part, 


excl. Zeyher 4017; N.E.Br. in J. Linn. Soc. 48: 25 (1928). Type: Promont. 
b. Spei, Oldenburg & Masson (BM, holo— ) - “Promont. b. Spei probably 
refers to the Cape in general. 

Gladiolus striatus Soland. ined. in Hb. Banks; non Jacq. (1789). 
Montbretia striata Soland. ex. Bak. in J. Bot. Lond. 14: 337 (1876) & in 
J. Linn. Soc. 16: 168 (1877). Type as for T. bakeri. 

Tritonia striata (Bak.) Klatt in Dur. & Schinz, Consp. FI. Cap. 5: 208 
(1895) in part, excl. syn. T. capensis & Gladiolus roseus; horn, illeg., non 
Ker-Gawler (1805) ( = Babiana striata (Jacq.) G. J. Lewis). 

Gladiolus longiflorus L.f. Suppl.Pl. 96 (1781) in part, excl. lectotype 
(=T. cooperi ); Thunb. Diss. Gladiolo 19 (1784) & Prodr. Cap. 8 (1794) in 
part, excl. syn. Berg. & Delar. & Thunberg Herb. spec. (3; non Jacq. (1789) 

nec Vahl (1805). /moo'j • 

Tritonia longiflora (L.f.) sensu N.E. Br. in J. Linn. Soc. 48: 25 (1928) in 

part, excl. Hb. Thunberg spec. (3; non Ker-Gawl. (1805) nec Ait. (1810). 

T. longituba R. C. Foster in Contr. Gray Herb. N. S. 114: 45 (1936) in 
part, excl. Hb. Thunb. spec. (3. 

Fig. 17. 


Flowers (50-)65-80(-90) mm long, cream or creamy-yellow. Perianth 
tube (30-)40-45(-55) mm long, ca. (l,5-)2 times the length of the segments, 


358 


Journal of South African Botany 


segments 18-25(-30) mm long, 7-10 mm wide, with the posticous segment 
up to 12 mm wide. Style (40-)50-65(-70) mm long. 

CAPE — 3321 (Ladismith): Seweweekspoort (-AD), Wall s.n., 14/11/38 (S), Wurts 
1221 (NBG); Elandsberg, Wurts 1190 (NBG); Southern side of Seweweekspoort, 
Delpierre s.n. sub De Vos 2436 (STE); Outside Waterkloof, Ladismith (-CB), Gillett 
1913 (BOL, STE), De Vos 2413 (STE); Hoekoe, De Vos 2414 (STE); Poort, Taylor 
8978 (STE); Vanwyksdorp, Compton 3878 (BOL), Archer 686 (BOL); 2 mis ENE of 
Vanwyksdorp, Acocks 20784 (PRE); Rooiberg Pass, ridge W of summit (-DA), Oliv- 
er 5291 (PRE, STE), Wurts 1653 (NBG); Rooiberg, Nicolaaskloof, Thompson 3349 
(STE). 

— 3322 (Oudtshoorn): Cango Caves (-AC), Oliver NBG 63028; N side of Outeniquas 
nr Moerasrivier (-CC), Esterhuysen 19488 (BOL); Zebra (-CD), Esterhuysen 19357 
(BOL); Outeniqua Pass, Loubser 2078 (NBG); Camfer, Esterhuysen 7106 (BOL); 4,2 
miles WNW of Camfer station, Acocks 23071 (PRE); Laudina (-DB), Esterhuysen 
6365 (BOL); 

— 3323 (Willowmore): Hills nr Avontuur (-CA), Fourcade 3166 (BOL, PRE, STE, 
MO), Marloth 10937 (PRE); Uniondale poort, Acocks 19975 (PRE); Foothills of 
Kammanassie Mts nr Avontuur, Esterhuysen 16479 (BOL); Prope Avontuur, Bolus 
2483 (K); Hills S of Haarlem (-CB), Fourcade 3496 (STE); Misgund hills (-CD), 
Esterhuysen 6951 (BOL), Fourcade 5477 (STE). 

— 3419 (Caledon): Zwellendam am Berge bei Puspasvalei (-BA), Ecklon & Zeyher 
Irid 96 (MO, not G). 

— 4322 (Mossel Bay) — 3322 (Oudtshoorn): Mossel Bay to George, Barker 7722 
(NBG). 

Without precise locality: Drege 8409 (G, L, K); Burman (G); Thunberg 1040 
(UPS) sub Gladiolus longiflorus. 


b. subspecies lilacina (L.Bol.) De Vos, stat. nov. 


T. lilacina L. Bol. in Ann. Bol. Herb. 2: 161 (1918). Type: Cape, exhibi- 
ted at the Cape Town wild flower show among plants sent from Riversdale, 
BOL 13755 (BOL, holo.-). 

Icon; Coloured drawing by M. Page in BOL. 

Flowers 35-50 mm long, pale mauvish-pink with fine purplish veins out- 
side on outer segments, or sometimes cream. Perianth tube 20-30 mm long, 
ca. 1,5 times the length of the segments; segments 15-22 x 5-8 mm, with 
the posticous segment up to 10 mm wide. Style 27-35(-40) mm long. 

CAPE — 3322 (Oudtshoorn): Swartberg Pass (-AC), Wall s.n., 13/11/38 (S); Cango 
valley mts (-AD), Marloth 7554 (PRE); Grootkruis, Thorns NBG 63029; 
Oudtshoorn (-CA), Yeoman NBG 1717/25 (BOL, NBG), Deas SAM 11640, Jenner 
252 (BOL). 

Without precise locality: Sent to Cape Town wild flower show, Oct. 1914, 1916; 
probably ex Oudtshoorn, BOL 13370, BOL 14683 (SAM); ex Riversdale, BOL 
13755. 


African genus Tritonia Ker-Gawler: part 2 359 

11. Tritonia pallida Ker-Gawl. in Curtis’s bot. Mag. sub t. 1275 (1810). 

Plants (150-)200-400(-600) mm long. Corm ovoid, 10-20 mm diam., 
tunic fibres fine, reticulate. Scape (150-)200-400(-600) mm long, often 
somewhat flexuose or curved, often with one or several branches, Leaves ca. 
5-6, linear-lanceolate, curved or suberect, acute or acuminate, 
(100-)150-300(-500) x 8— 15(— 25) mm, with a slender middle vein, reaching 
to the base of the spike; cauline leaves 2-3, subulate, 15— 30(— 50) mm long. 
Spike distichous, simple or with 1 or more branches, rather lax, with 4-8 or 
more flowers. Bracts membranous, ovate or ovate-lanceolate, straw- 
coloured, (7-)12-20 mm long; outer minutely 3-toothed or sometimes 2- 
toothed or irregularly toothed, veined, finely sphacelate upwards; inner bi- 
dentate or shortly bifid, with wide papery margins, 2-veined, sometimes 
slightly shorter than the outer. Flowers zygomorphic, (40-)55-80(-100) mm 
long, suberect, white or cream or sometimes pale lilac, greenish-yellow in 
the throat, with purple veins on the tube. Perianth tube (25— )35— 65(— 75) 
mm long, tubular, slightly curved, widened in the upper quarter to 5-8 mm 
diam., often slightly pouched on the anticous side; segments unequal, ob- 
tuse, (8-)10-17(-20) mm long, median posticous segment erect, obovate or 
suborbicular, (7— )10— 13 mm wide, lateral posticous segments spreading, ob- 
long, 4-8 mm wide, anticous segments spreading, oblanceolate-obovate or 
oblong, 4-6 mm wide, often with a long, low yellow-green excrescence (cal- 
lus) in the throat. Stamens curved against the posticous segment; filaments 
(7—) 10—15 mm long, white; anthers 5-8 mm long, sometimes with curved 
tips, mauve to purple, reaching halfway up the perianth segments, pollen 
purple. Style (32-)50-75(-85) mm long; stigmatic branches 2-4 mm long, 
reaching the anther tips or higher. Capsules 12-20 mm long, ellipsoid; seeds 
brown, shiny with a reticulate-foveate testa. Chromosome number 2n = 22. 

Flowering period: September to November. 

Distribution: Little Karoo from Montagu to Ladismith, and coastal dis- 
tricts from Caledon to Riversdale (Fig. 15). 

This species is distinguished by its pale long-tubed flowers, often with a 
low, greenish-yellow callus on the median or on all three anticous perianth 
segments, and by linear-lanceolate, often somewhat spreading leaves. The 
typical subspecies can easily be confused with T. flabellifolia which has al- 
most similar flowers and habit. T. pallida differs in its shorter outer bracts 
which are three-toothed or irregular at the top; the perianth usually has 
small calli and the leaves are generally wider and, when dry, not as brown as 
in T. flabellifolia. 

The small median excrescences (calli) in the throat on the anticous seg 
ments vary in number, size and shape, from a low median ridge on each an 


360 


Journal of South African Botany 



African genus Tritonia Ker-Gawler: part 2 


361 


ticous segment to a semi-orbicular (fide L. Bolus, 1926) callus on the median 
segment and a low callus (or none) on the two lateral anticous segments. 
These calli are not obvious in dried material. 

Jacquin (1789) figured two long-tubed Iridaceous species as Gladiolus 
longiflorus (Ic. t. 262 and 263), erroneously identifying them with G. longi- 
florus L.f. and G. longiflorus Thunb., t. 262 being the present species and 
t. 263 Ixia paniculata. Willdenow (1797) transferred both to Ixia as I. longi- 
flora , citing Jacquin’s t. 262 under his var. a which further comprised several 
other species. Vahl (1805) also united the two taxa, citing Jacquin’s t. 262 as 
his var. (3 of Gladiolus longiflorus. 

Ker-Gawler (1810), on transferring Jacquin’s G. longiflorus (Ic. t. 262) 
to Tritonia. could not use the combination T. longiflora for this species, as 
he had already incorrectly used the name in 1805 for Ixia paniculata. He 
therefore named it T. pallida, citing Jacquin, as well as Willdenow’s var. a 
and Vahl’s var. (3. Baker, in his treatment of T. pallida (1877, under Mont- 
bretia; 1892, 1896) correctly excluded G. longiflorus L.f. from the citations 
as he realised that Jacquin’s and the younger Linnaeus’s G. longiflorus were 
not the same species. 


Key to the Subspecies 

Flowers longer than 55 mm, with the perianth tube three to four times as long as. 

the segments a - subsp. pallida 

Flowers shorter than 50(— 55) mm, the perianth tube ca. twice (or less) as long as 
the segments b. subsp. taylorae 

a. subspecies pallida 


Tritonia pallida Ker-Gawl. in Curtis’s bot. Mag. sub t. 1275 (1810), Bak. 
Handb. Irid. 192 (1892) & in FI. Cap. 6: 122 (1896); Klatt in Abh. naturf. 
Ges. Halle 15 : 358 (1882) & in Dur. & Schinz, Consp. FI. Afr. 5: 206 
(1895). Iconotype: Prom. b. Spei, Jacquin Ic. t. 262 (1789) sub Gladiolus 
longiflorus (lecto-). 

Montbretia pallida (Ker-Gawl.) Bak. in J. Linn. Soc. 16: 168 (1877). 
Gladiolus longiflorus L.f. sensu Jacq. Ic. t. 262 (1789) & Coll. 5: 23 
(1796), excl. syn.; non L.f., nec. Thunb. 

G. longiflorus var. (3 Vahl, Enum 2: 111 (1805). 

leones: Jacq. Ic. t. 262; Lewis, ined., in BOL; this work Fig. 18. 


Tritonia pallida: a, plant 
bract; c, flower half in a 



362 


Journal of South African Botany 


Bracts (8-)12-20 mm long. Flowers (55-)70-90(-100) mm long, white or 
cream, slightly greenish-yellow in the throat. Perianth tube (45-)55-65(-75) 
mm long, 3-4 times longer than the segments; segments sometimes with a 
small callus on one, or all three, anticous segments. Filaments (10)13-15 
mm; anthers 7-8 mm, purple. Style (55-)60-75(-85) mm long. 


CAPE — 3319 (Worcester): Prope Robertson (-DD), Marloth 8001 (PRE). 

— 3320 (Montagu): Montagu Division, Compton 3887 (BOL, NBG, PRE); Laings- 
burg, Whitehill (-BA), Barker 4881 (NBG); Montagu (-CC), Taylor 4034 (NBG), 
Archer 536 (BOL); Fonteinkloof, Montagu, Lewis 4459 (SAM, NBG); Beyond Kog- 
manskloof, Middlemost SAM 68460\ Wildehondskloof, Barrydale side, Goldblatt 
2845 (MO); Barrydale (-DC), Hafstrom & Acocks 281 (PRE), Galpin 4708 (PRE, 
GRA), Sorensen 315 (C). 

— 3321 (Ladismith): Vleiland, N of Klein Swartberge (-AC), Thompson 3182 (PRE, 
STE); Between Laingsburg & Ladismith, Loubser 2254 (STE); Ladismith (-AD), 
Bayliss 2436 (NBG); 9 mis W of Ladismith, Salter 6326 (BOL); 3,4 mis from Ladi- 
smith to Barrydale (-CB), Marsh 1425 (PRE, STE, S), De Vos 2412 (STE); 7 mis E 
by S of Ladismith, Acocks 20629 (NBG); Vanwyksdorp (-CD), Van Breda 748 
(PRE), Archer 687 (BOL, NBG); Calitzdorp (-DA), Van Breda 748 (PRE); Dwars- 
in-die Weg, Rooiberg Pass, Thompson 1431 (PRE, STE); Between Cloete’s Pass and 
Wagenboom, Goldblatt 4167 (MO). 

—3419 (Caledon): Slopes above Greyton (-BA), Jones s.n., Nov (STE). 

— 3420 (Bredasdorp): Swellendam (-AB), Archer 535 (BOL). 

— 3421 (Riversdale): Waterval, Riversdale (-AB), Muir 3750 (BOL). 


b. subspecies taylorae (L. Bol.) De Vos, stat. nov. 

Tritonia taylorae L. Bol. in Ann. Bol. Herb. 4: 44 (1926). Type: Cape, 
prope Oudtshoorn, in collibus aridis Bankhoogte, Taylor 251 (= BOL 
12347) (BOL, holo-). 

Icon: M. Page, ined. in BOL (flower slightly enlarged). 


Bracts 5— 10(— 13) mm long. Flowers 40-45(-58) mm long, pale lilac or 
cream, sometimes richly veined. Perianth tube (25-)30-35(-40) mm long, 
ca. twice or sometimes 1,5 times the length of the segments; anticous seg- 
ment sometimes with a semi-orbicular process (f. L. Bolus). Filaments 7-10 
mm; anthers 4-6 mm long. Style 32-40(-48) mm long. 

CAPE — 3319 (Worcester): Slopes of Rabiesberg (-DA), Lewis s.n., 26/9/ 35 (BOL); 
Dassieshoek Pass (-DD), NBG 91317. 

— 3320 (Montagu): Montagu Baths (-CC), Martley BOL 31767, Page BOL 31778; 
Near Wild Flower Garden, Hall s.n., 21/9/58 (NBG), De Vos 2411 (STE). 

— 3321 (Ladismith): Along Ladismith road, 10 mis out, Lewis BOL 21648 (PRE); 
Gouritz River valley beyond Cloete’s Pass (-DC), Stayner s.n., 20/10/62 (NBG); 
Cloete’s Pass (-DC/-D), Muir 2392 (BOL). 

— 3322 (Oudtshoorn): Bankhoogte pr. Oudtshoorn (-CA), Taylor 251 (BOL). 
Without precise locality: Thunberg sub Freesea rosea (S). 


African genus Tritonia Ker-Gawler: part 2 


363 


12. Tritonia flabeliifolia (Delaroche) G.J. Lewis in J1 S. Afr. Bot 7- 30 
(1941). 

Plants (200-)250-400(-600) mm long. Corm depressed globose, 
20-30(-40) mm diam., tunic fibres fine, somewhat reticulate, almost parallel 
towards the base. Scape (200-)250-400(-600) mm long, straight or slightly 
curved, simple or with 1-3 branches. Leaves 6-9, linear-lanceolate or linear, 
acuminate, suberect, 100-250(-450) x (2-)4-6(-10) mm, reaching to the 
base of the spikes, drying to a golden-brown or nut-brown, with a pale mid- 
dle vein; cauline leaves few, subulate. Spikes 1-3, secund, dense or lax, with 
3-7 flowers in a spike. Bracts lanceolate with elongated attenuate tips, 
(10-)15-30(-40) mm long; outer acuminate or attenuate or rarely with three 
acuminate tips close together, with a strong median vein towards the tip, 
greenish near the base, elongating to 40-70 mm after flowering; inner biden- 
tate with acuminate teeth, 2-veined, shorter than the outer bract. Flowers 
zygomorphic, (42-)65-85(-95) mm long, suberect, white, cream or pale 
pink, with a red stripe or a narrow yellow, red outlined stripe or blotch on 
the three anticous segments, and often magenta in the throat. Perianth tube 
(25-)45-60(-65) mm long, tubular, widened gradually in the upper quarter 
to ca. 10 mm diam., (1 ,5— )2— 3 times the length of the segments; segments 
unequal, obtuse, (12— )16— 25 mm long, the posticous segment obovate, 
10-15 mm wide, the other segments obovate-oblanceolate, (5-)7-10 mm 
wide. Stamens curved towards the posticous segment; filaments (13— ) 15— 18 
mm; anthers 5-7(-9) mm long, violet, reaching less than halfway to halfway 
or higher up the perianth segments, pollen pale. Style (30-)55-75(-85) mm 
long; style branches 4-7 mm long, overtopping the anthers and sometimes 
reaching the tips of the perianth segments. Capsules obovoid, 7-10 mm 
long; seeds shiny brown, 1,5 mm diam. Chromosome number 2n = 44. 

Flowering period: late September to early December. 

Distribution: Cape south coastal districts from Caledon to Riversdale, 
and to Worcester (Fig. 15). 

This species is readily distinguished by its long-tubed zygomorphic flow- 
ers that have no calli in the throat, its long, acuminate bracts which elongate 
considerably after flowering, and by linear or linear-lanceolate leaves which 
become brown (RHS 174B) when dry and have a slender pale middle vein. 
The perianth tube is two to three times longer than the segments (or rarely 
only 1,5 times longer), and the largely membranous bract is longer than the 
bracteole and somewhat green at its base; it is usually acuminate or rarely, 
in subspecies thomasiae, three-toothed, and is often slightly keeL^ upwards. 
In habit and floral characters it resembles T. pallida closely, but is distin- 
guished by its bracts. 


364 


Journal of South African Botany 



Fig. 19 

Tritonia flabellifolia var. major: a, plant and front view of flower; b, outer (left) and 
inner (right) bract; c, flower half in a median cut; ca, low ridge-like callus. ( De Vos 

2437). 

The species has a long and complicated history. D. Delaroche (1766) de- 
scribed it in his thesis as Ixia flabellifolia. N. L. Burman (1768) had no clear 
idea of Delaroche's species, as he noted under this name that it varied in its 
whitish and blue flowers; he also described 1. tubulosa, with two varieties, 
one of which is the present species. Jacquin (1796) named the species Gladi- 


African genus Tritonia Ker-Gawler: part 2 


365 


olus roseus and Aiton (1810), adopting Jacquin’s epithet, named it Tritonia 
rosea. (Klatt later described a different species from the eastern Cape Prov- 
ince as T. rosea - a later homonym.) 

Ker-Gawler (1803, and again 1813) illustrated and described the plant as 
T. capensis, basing it nomenclaturally on Houttuynia capensis Houtt. (1780). 
Figure 85/3 accompanying Houttuyn's description should be chosen as lecto- 
type, as Houttuyn mentioned that the plant he described was lost. This 
figure shows a regular flower and can therefore not be present Tritonia 
species. 

Bentham and Hooker (1883) were the first to cast doubt on Ker-Gawler’s 
equating the present species with Houttuynia capensis, observing (p.706) 
that the shape of the perianth in Houttuyn’s figure agrees badly with 
T. capensis, and that the short spathe fits neither Tritonia nor Acidanthera. 
E. D. Merrill (1938) pointed out Ker-Gawler’s error (J. Arnold Arbor. 19: 
328) and stated that the Houttuyn figure seemed “manifestly to appertain to 
Ixia Lewis (1962) identified Houttuyn’s plant with its symmetrical flower 
and equal perianth segments, with Ixia paniculata Delaroche, a species that 
occurs in several forms. 

Neither Klatt nor Baker realised Ker-Gawler’s misapplication of the 
name: Klatt (1863) called the present species T. capensis and Baker (1877) 
named it Montbretia capensis, after Voigt’s transference of it to Montbretia 
(1845). Later (1892, 1896) Baker transferred the species to Acidanthera, as 
A. capensis. Lewis, in her revision of Acidanthera (1941), transferred the 
species to Tritonia, as T. flabellifolia Delaroche, this being its oldest epithet. 
She observed that although the bract is slightly larger and more acute than is 
usual in Tritonia, the corm, leaf and floral characters, as well as the mem- 
branous bract, belong to this genus. 

Ecklon (1827) named the species Freesea secunda, an illegitimate name 
as he published no description of the genus (Nordenstam, 1972). 

Ker-Gawler (1813) described two varieties of his T. capensis in Curtis’s 
Botanical Magazine, namely var. major figured as t. 618 and var. minor 
(t. 1531). They differ in flower size and colouring and in the marks on the 
perianth, var. major being pale pink with a red margined yellow blotch on 
each anticous segment; var. minor has cream flowers with a narrow red 
stripe on these segments. Var. major varies slightly, as is to be seen in Jac- 
quin’s leones t. 261 (1789) which shows a pale pink flower with two red lines 
on either side of a median yellow line on the anticous segments. Voigt 
(1845) regarded the two varieties as separate species, naming t. 618 Mont- 
bretia rosea and t. 1531 M. capensis. 

As the greater majority of herbarium specimens can readily be separate 
into the two varieties established by Ker-Gawler (1813), his varieties arc 
upheld. The Burman specimen, as well as the specimens in the Van Ro /en 


366 


Journal of South African Botany 


herbarium amongst which Delaroche’s type material probably occurs, be- 
longs to Ker-Gawler’s var. minor , and this therefore becomes the typical 
variety. 

A few recent collections by Thomas from the Bot River-Hawston area 
have flowers with shorter perianth tubes and bracts which are either acumi- 
nate or have three acute teeth very close together. They are regarded as a 
third variety. 

Key to the Varieties 

1 Perianth tube 2-3 times longer than the segments: 

2. Flowers less than 50 mm long, rarely up to 60 mm; perianth tube 25-30( — 40) 

mm long a. var. flabellifolia 

2. Flowers (60— )65— 75(— 95) mm long; perianth tube (45— )50— 65 mm long 

b. var. major 

1. Perianth tube 1,5 times longer than the segments c. var. thomasiae 

a. var. flabellifolia 

Ixia flabellifolia Delaroche Descr. pi. aliq. nov. 24 (1766). Type: s. loc., 
Van Royen (L, neo-). But see below. 

Acidanthera flabellifolia (Delaroche) N.E.Br. in Kew Bull. 1929: 135 
(1929). 

Tritonia flabellifolia (Delaroche) G. J. Lewis in J1 S. Afr. Bot. 7: 30 
(1941); Goldbl. & Barnard in J1 S. Afr. Bot. 36: 310 (1970). 

T. capensis (Houtt.) Ker-Gawl. var. minor Ker-Gawl. in Curtis’s bot. 
Mag. 37 t. 1531 (1813) in part, excl. basionym Houttuynia capensis and 
Gladiolus longiflorus. Iconotype: Curtis’s bot. Mag. t. 1531 (lecto-). 

T. capensis (Houtt.) Ker-Gawl. sensu Klatt in Linnaea 32 : 757 (1863), 
excl. syn. Gladiolus and Bot. Mag. t. 618; & in Dur. & Schinz, Consp. FI. Afr. 
5 : 204 (1895), excl. basionym. 

Montbretia capensis (Houtt.) Voigt, Hort. Suburb. Calcut. 611 (1845); 
Bak. in J. Linn. Soc. 16: 168 (1877), excl. basionym and Bot. Mag. t. 618. 

Acidanthera capensis (Houtt.) Bak. Handb. Irid. 187 (1892) & in FI. 
Cap. 6: 133 (1896), excl. cit. Bot. Mag. t. 618 and syn. Gladiolus. 

Outer bracts 8-15 mm long, acute to acuminate. Flowers 42-45(-60) mm 
long, white, cream or pale pink, the anticous segments with a narrow median 
red stripe or yellow stripe outlined in red in the throat. Perianth tube 
25-30(^40) mm long. Style 30-35(-45) mm. 

CAPE — 3319 (Worcester): Between Robertson and McGregor (-DD), Leipoldt 3602 
(BOL, PRE). 

— 3320 (Montagu) to — 3419 (Caledon): Swellendam bei Berge Puspasvalei Voor- 
mansbosch, etc. — 3419-BA — 3320-DC), Zeyher 4017 (S). 


African genus Tritonia Ker-Gawler: part 2 


367 


—3419 (Caledon): Genadendal (-BA), Prior s.n. (K); Greyton, Jones s.n. (STE). 

Without locality: Van Royen collection (L); Barman sub Ixia flabellifolia (G). 

Type: N. E. Brown (1929) found four specimens in the Burman collec- 
tion at Geneva labelled Ixia flabellifolia Laroche. He identified three as Ba- 
biana and the fourth as the true I. flabellifolia , as it “so accurately agrees 
... in all characters except the length of the style” with Delaroche’s de- 
scription of 1766. He indicated this specimen as type, as he believed the 
label on this sheet to be in the handwriting of D. Delaroche. On my en- 
quiry, Dr. H. M. Burdet of Geneva, specialist in old handwriting, was un- 
able to confirm that this label was indeed in Delaroche’s hand. 

Goldblatt and Barnard (1970), in their work on the Iridaceae of Daniel 
Delaroche’s thesis, state that neither the Burman specimen nor those they 
found in the Van Royen herbarium in Leiden, have any claim to be the type 
material. They therefore chose a neotype from amongst the Van Royen 
specimens, namely one labelled Tritonia viridi Ker proximo and Ixia flabelli- 
folia La Roche (in two different handwritings). 

If, however, the label on Burman’s specimen which N. E. Brown iden- 
tified as type, is indeed found to be in D. Delaroche’s hand, then this speci- 
men must be the lectotype which will supersede the neotype indicated by 
Goldblatt and Barnard. 

b. var. major (Ker-Gawl.) De Vos, comb. nov. 

Tritonia capensis (Houtt.) Ker-Gawl. sensu Ker-Gawl. var. major Ker- 
Gawl. in Curtis’s bot. Mag. sub t. 1531 (1813) & Bot. Mag. t. 618 (1803) in 
part, excl. basionym Houttuynia capensis, and Gladiolus striatus ; & in Kon. 
& Simms, Ann. bot. 1: 228 (1805). Iconotype: Curtis’s bot. Mag. t. 618 
(lecto-). 

T. capensis (Houtt.) Ker-Gawl. sensu Klatt in Linnaea 32: 757 (1863) in 
part, excl. syn. Gladiolus and Bot. Mag. t. 1531; & in Abh. naturf. Ges. 
Halle 15 : 358 (1882). 

Montbretia capensis (Houtt.) Voigt sensu Bak. in J. Linn. Soc. 16. 168 
(1877) in part, excl. basionym H. capensis, and Bot. Mag. t. 1531. 

Acidanthera capensis (Houtt.) Bak. Handb. Irid. 187 (1892) & in FI. 
Cap. 6: 133 (1896), excl. Bot. Mag. t. 1531. 

Gladiolus roseus Jacq. Ic. Rar. t. 261 (1789) & Coll. 5: 22 (1796). Icono- 
type: Ic. t. 261 (lecto-). 

Tritonia rosea (Jacq.) Ait. Hort. Kew. ed. 2, 1: 91 (1810), excl. syn. 
Houttuynia-, non Klatt (1863, 1895), nec Bak. (1892, 1896). 

Montbretia rosea (Jacq.) Voigt, Hort. Suburb. Calcut. 610 (1845), non 
Bak. 1877 (= T. disticha var. rubrolucens) . 


368 


Journal of South African Botany 


Freesea secunda Eckl. Top. Verz. 30 (1827), nom. illeg. , gen. non descr.; 
Nordenstam in J1 S. Afr. Bot. 38: 296 (1972). 

Tritonia secunda (Eckl.) Steud. Nom. Bot. ed. 2, 2: 719 (1841) nom. 
nud. 

Ixia tubulosa N. L. Burm. Prodr. FI. Cap. 1 (1768) in part, excl. lecto- 
type (= Babiana tubulosa)-, N.E. Br. in Kew Bull. 1929: 137. See below, 
leones: Curtis’s bot. Mag. t. 618; Jacq. Ic. Rar. t. 261; this work Fig. 19. 


Outer bracts (15-)20-40 mm long, acuminate or attenuate. Flowers 
(60-)65-75(-95) mm long, pale pink or white, with a yellow median blotch 
or stripe often outlined in red on the three anticous perianth segments. Per- 
ianth tube (45-)50-60(-65) mm long. Style (55-)60-85) mm. 

CAPE — 3319 (Worcester): Villiersdorp (-CD), Barnard BOL 31782 ; McGregor, foot 
of Boesmanskloof Pass (-DD), Lewis 6203 (NBG, MO); Robertson, Flail 719 
(NBG); Langvlei, Robertson, Van Breda 2063 (PRE). 

— 3419 (Caledon): Caledon (-AB), Guthrie BOL 31781 ; Between Caledon and 
Riviersonderend (-BA), Barker 6843 (NBG); Am Baviaansberg bei Genadenthal, 
Ecklon 282 (S); Riviersondereinde, Compton 21795 (NBG), Zeyher 4017 (SAM, K), 
Pappe SAM 21113, 21114- Near Elim (-DA), Frowein PRE 16119, Bolus BOL 
31780\ Between Vogelvlei and Wisedrift (-DB), Mathews 32 (NBG). 

— 3420 (Bredasdorp): Dasberg near Stormsvlei (-AA), Stokoe SAM 59773; Bonte- 
bokpark (-AB), Marais 1 (PRE); 13 mis from Port Beaufort towards Malgas (-BC), 
Lewis 6127 (NBG); Potteberg, David 292 (NBG); Near Cape Infanta (-BD), Ester- 
huysen 29358 (BOL, MO); Near Bredasdorp (-CA), Hafstrom & Acocks 2069 (S). 

— 3421 (Riversdale): Corenterivier farm (-AA), Muir 5363 (BOL, PRE, GRA); In 
lapidosis pr. Riversdale (-AB), Schlechter 1815 (PRE, G, K); 5 mis S of Riversdale 
on Blombos road, Thomas s.n. (NBG); Near Zoetmelks River, Burchell 6739 (K); 
Around Albertinia (-BA), Muir 1157 (BOL). 

Without precise locality: lrid. 282 (G); Burman (G) sub Ixia tubulosa. 


Burman’s description (1768) of Ixia tubulosa is that of two species, and 
three of the five specimens so named in his collection in Geneva are Babiana 
tubulosa (Burman f.) Ker-Gawler. One of these has been chosen as the type 
of this name (Lewis, 1959). The other two specimens in the collection are 
Tritonia flabellifolia and Engysiphon pictus (Foster) G.J. Lewis. 

c. var. thomasiae De Vos, var. nov. 

Ab aliis varietatibus Tritoniae flabellifoliae distinguitur tubo perigonii 
non nisi 1,5-plo longiore quam segmentis, bractea exteriore aliquando tri- 
dentata ad apice, dentibus approximatis, acuminatis. 

Type: Cape, Caledon, Protea farm between Hawston and Bot River, Thomas 
s.n. NBG 77202 (NBG, holo-, STE). 

Outer bracts 10-25 mm long, acuminate or sometimes with three long 


African genus Tritonia Ker-Gawler: part 2 


369 


acuminate tips close together, keeled near the tip. Flower 50-67 mm long, 
rose-pink inside, with a yellow blotch outlined in red on the anticous peri- 
anth segments, deeper pink outside. Perianth tube 30-45 mm long, only ca. 
1,5 times the length of the segments. Style 45-55 mm. 

CAPE — 3420 (Caledon): Protea farm, between Hawston and Bot River (-AC), 
Thomas NBG 70.508, 77.202, 87.945, Thomas s.n., 6/11/64 (STE). 


13. Tritonia lineata (Salisb.) Ker-Gawl. in Kon. & Sims, Ann. Bot. 1: 228 
(1805). 

Plants (150-)400-600(-850) mm long. Corm shortly ovoid, 10-20 mm 
diam., sometimes with small cormlet offsets at the base; tunic fibres fine, re- 
ticulate, towards the base long-reticulate. Scape simple or branched, (150-) 
400-600(-850) mm long. Leaves 4-8, linear-lanceolate or linear-ensiform, 
(sub)erect or spreading, acute, (40-)150-300(-500) x (2— )7— 10(— 15) mm, 
with a prominent vein near each edge and in the middle, usually not reach- 
ing the base of the spike, often drying to brown or dark brown with pale 
veins; cauline leaves 1-3, 25-100 mm long. Spike somewhat secund, lax, 
sometimes with one branch, with (1— 2— )7— 12 flowers. Bracts membranous, 
usually brown especially in the upper half, 7— 10(— 12) mm long; outer 3-5- 
toothed with the median tooth sometimes reduced, or irregularly toothed, 
striate, with a stronger median vein; inner bifid with acuminate teeth, with 2 
dark veins and papery margins. Flowers slightly zygomorphic, funnel- 
shaped, (25-)30-35(-40) mm long, cream, pale yellow or straw-coloured to 
pale apricot, with dark veining and sometimes apricot tints outside espec- 
ially towards the tips of the tepals. Perianth tube funnel-shaped, 10-15 mm 
long, narrow in the lower half, widened to 8 mm diam. upwards; segments 
obtuse, subequal in length, (15— )18— 20(— 23) mm long, the median posticous 
segment (8— )10— 12(— 14) mm wide, obovate-lanceolate, the other segments 
5-8(10) mm wide, oblanceolate, sometimes with a low yellow, ridge-like cal- 
lus on the median anticous segment. Stamens curved towards the posticous 
side; filaments 7— 10(— 12) mm; anthers unilateral, 5-9 mm long, pale or with 
dark lateral stripes, reaching halfway or almost halfway up the perianth seg- 
ments, pollen cream. Style (15—) 18—25 mm; stigmatic branches 3-5 mm 
long, reaching or overtopping the anther tips. Capsules subrotund-trigonous 
or shortly obovoid-trigonous, ca. 8 mm long; seeds dark brown, angled, 2,5 
mm diam. 

Flowering period: August to November (to December). 

Distribution: eastern and south eastern Cape Province, Transkei, Natal, 
north eastern Orange Free State, south eastern Transvaal, and Lesotho 
(Fig. 16). 


370 


Journal of South African Botany 


This species was first described by Aiton (1789) as var. stricta of Ixia 
squalida. Seven years later Salisbury, realising its difference, described it as 
Gladiolus lineatus and cited Aiton’s variety. He was followed by Ker-Gawler 
(1800) who transferred it in 1805 to his new genus Tritonia. 

The species is distinguished by its short-tubed cream, straw-coloured, 
yellow or salmon-pink flowers (e.g. RHS 20B, C, 28C, D, 11B, C), “finely 
and curiously pencilled with dark lines” (Ker-Gawler, 1800), and by leaves 
which show three prominent veins, two of which are near the leaf margins, 
and especially prominent in the dried state. A small yellow area occurs in 
the throat on the anticous perianth segments of the cream and straw- 
coloured flowers, and sometimes the median anticous segment has a low cal- 
lus in the form of a longitudinal ridge. It differs from the closely related 
T. disticha in flower colouring and slightly shorter perianth tube, as well as 
in its different flowering period. 

Specimens with cream or straw-coloured flowers, often with orangy tints 
on the reverse, occur in the eastern Cape Province and Natal. Such are 
figured in Curtis’s Botanical Magazine t. 487 and in Redoute’s Liliacees 
t. 55. In herbarium specimens the tepal tips often become pale brown. 

Most specimens from the Orange Free State, Lesotho and the Drakens- 
berg region of Natal have pale yellow flowers without orange tints outside, 
and often with less distinct veining. This form was described as T. flavida by 
Schlechter (1908). Another variation which has pale pink flowers, was de- 
scribed as T. kraussii by Baker (1892). These are apparently colour varia- 
tions only and are not upheld here as separate varieties. 

The two specimens which Redoute figured (t. 55 and 400) were de- 
scribed as var. minor and var. major respectively ( Liliacees sub t. 400, 
1813). There is not enough evidence to uphold these varieties as they are 
connected by numerous intermediates. But certain collections from the 
south eastern Cape Province comprise distinctly smaller plants with much 
shorter and narrower, erect leaves, and these are now described as a separ- 
ate variety. 

Key to the Varieties 

Plants more than 300 mm long; leaves more than 150 X 5 mm, suberect or spread- 
ing a. var. lineata 

Plants up to 300 mm long; leaves less than 150 x 5 mm, usually 
erect b. var. parvifolia 

a. var. lineata 

Tritonia lineata (Salisb.) Ker-Gawl. in Kon. & Sims, Ann. Bot. 1: 228 
(1805); Ait. Hort. Kew. ed. 2, 1: 91 (1810); Klatt in Linnaea 32: 759 (1863) 
& in Dur. & Schinz, Consp. FI. Afr. 5: 206 (1895); Bak. Handb. Irid. 194 
(1892) & in FI. Cap. 6: 25 (1896). 


African genus Tritonia Ker-Gawler: part 2 


371 


Gladiolus lineatus Salisb. Prodr. 40 (1796); Ker-Gawl. in Curtis’s bot. 
Mag. t. 487 (1800). Syntypes: sub Ixia squalida var. stricta, Hort. Kew. 1781 
(BM, lecto-) & 1778 (BM). See below. 

G. lineatus Salisb. var. minor & var. major , Red. Lil. 7 t. 400 (1813). 
Iconotypes: Red. Lil. t. 55 & t. 400 respectively. 

Montbretia lineata (Salisb.) Bak. in J. Linn. Soc. 16: 169 (1877), excl. 
syn. Ixia thunbergii. 

Tritonixia lineata (Salisb.) Klatt in Abh. naturf. Ges. Halle 15: 357 
(1882). 

Ixia squalida Ait. var. stricta Ait. Hort. Kew. 1: 61 (1789). Syntypes: 
Hort. Kew. 1778 (BM), 1781 (BM, lecto-). 

Gladiolus venosus Willd. Enum. 1: 58 (1809). Type: Cap. b. Spei, Will- 
denow 929 (B, holo-). 

Tritonia catenularis Salisb. in Roy. Hort. Soc. Trans. 1: 319 (1812), nom. 
superfl. Type as for G. lineatus. 

Ixia reticulata Thunb. FI. Cap. ed. 2, 60 (1823). Type: Cap. b. Spei, 
Thunberg 989 (UPS, holo-). 

Tritonia kraussii Bak. Handb. Irid. 194 (1892) & in FI. Cap. 6: 125 
(1896), excl. syn. Type: Natal, summit of Table Mountain, Krauss 200 (K, 
lecto-). 

T. flavida Schlecht. in Bot Jb. 40: 91 (1908). Type: Jolivet, unweit Fair- 
field, Alexandra Country, Rudatis 91 (B, holo-). 

leones: Curtis’s bot. Mag. t. 487; Red. Lil. 1 t. 55 & 7 t. 400; this work 

Fig. 20. 

Type: As Salisbury (1796) cited only Ixia squalida |3 Soland. in Ait., the 
type of the latter in BM is also the type of T. lineata. There are two sheets 
of type material in BM dated 1778 and 1781. The latter, being the better 
specimen, is chosen as lectotype. 

Plants (300-)400-600(-850) mm long. Scape (300-)400-600(-850) mm, 
simple or branched, with 7 or more flowers. Leaves linear-lanceolate or 
linear-ensiform, suberect or spreading, 150-300(-500) x (5— )7— 10(— 15) 
mm, usually not reaching the base of the spike, often drying to brown or 
dark brown with pale veins. Style up to 25 mm. Chromosome number 2n — 44. 

Distribution: widespread, from Uniondale and Humansdorp throughout 
the eastern Cape, Transkei, Natal, north eastern Orange Free State, to 
south eastern Transvaal and into Lesotho. 

TRANSVAAL— 2729 (Volksrust): Volksrust (-BD), Sidey 3973 (PRE, S). 

ORANGE FREE STATE— 2728 (Frankfort): Near Warden (-DD), Hafstrom & 
Acocks 332 (PRE). 


372 


Journal of South African Botany 



African genus Tritonia Ker-Gawler: part 2 


373 


—2827 (Senekal): Near Meguathling, Clocolan (-DC), Kinges 1216 (PRE); Ficks- 
burg (-DD), Rogers 15923 (BOL, GRA, Z), Galpin 13842 (PRE, S), Galpin BOL 


—2828 (Bethlehem): Signal Hill, Bethlehem (-AB), Potgieter 216 (PRE); Road 
Bethlehem to Clarens (-AD), Van Zinderen-Bakker 198 (PRE, BLFU); Golden 
Gate Glen Reenen (-DA), Liebenberg 7290 (PRE). 

—2829 (Harrismith): Platberg (-AC), Jacobs 2272, 3043 (PRE); Harrismith, Sankey 
292 (K); Kerkenberg, Van Zinderen-Bakker 1270 (BLFU). 

NATAL — 2730 (Vryheid): Paardeplaats 10 mis E of Wakkerstroom (-AD), De- 
venish 103 (PRE). 

— 2828 (Bethlehem): Royal National Park, Drakensberg (-DB), Sidey 2023 (S), 
Oliver 348 (NH); Mont-aux-Sources (-DD), Guthrie 4927 (BOL). 

— 2829 (Harrismith): Cathedral area near hotel (-CC), Schelpe 834 (NU). 

— 2830 (Dundee): Airfield (-AA), Shirley 126 (NU); Nkandla, 1 ml W of P.O. 
Qudeni (-DB), Codd 6999 (NH, NU). 

— 2831 (Nkandla): Intunzini (-BC), Garland NH 38636; Eshowe (-CD), Gerstner 
2551 (NH), Lawn 167 (NH); Umhlatuzi valley near Eshowe, Gerstner NH 22562; 
Melmoth (-CB), Porter s.n., Sept. 1955 (NBG, STE). 

— 2929 (Underberg): Impendhle distr, Loteni Nature Reserve (-AD), Phelan 106 
(NU); Kamberg (-BC), Wright 1538 (NU); Sani Pass (-CB), Gillies 113 (NU). 

— 2930 (Pietermaritzburg): Howick (-AC), Moll 1003 (NU); Manderston (-CB), 
Thomas 21 (NU); Camperdown (-DA), Barker 5310 (NBG); Cato Ridge, McLean & 
Ogilvie NH 27888; Inchanga, Marloth 4091 (K); Umzinyati (-DB), Wood 11569 (S); 
Inanda, Wood 669 (STE, K); Botha’s Hill (-DC), Hutchinson e.a. 45 (NH), Hutchin- 
son 4693 (K); Krantzkloof (-DD), Haygarth 68 (STE); Pinetown, Kloof, Johnson 
1348 ( NH). 

— 2931 (Stanger): Phoenix Umgebung (-CA), Schlechter 3114 (Z); Chatsworth 
(-CC), Coleman 178 (NH). 

— 3029 (Kokstad): Circa Kokstad (-CB), Tyson 507 (BOL, SAM, K), 1105 (BOL, 
SAM, etc.). 

—3030 (Port Shepstone): St. Michael (-AB), Thode 5098 (STE); Hlokozi, Alexandra 
City (-AD), Rudatis STE 2371; Jolivet unweit Fairfield, Rudatis 91 (B); Mvenyam 
(-CA), Bandert 9 (GRA); Shelley Beach (-CD), Strey 9241 (NU, SRGH). 

— 3130 (Port Edward): Ca 5 mis from Port Edward (-AA), Hilliard 1659A (NU). 

TRANSKEI— 3127 (Lady Frere): Near Cala (-DA), Royffe 174 (GRA). 

— 3128 (Umtata): Pomona, Ugie (-AA), Gill 143 (MO). 

—3129 (Port St Johns): Ntsubane Forest Sta. (-BC), Venter & Vorster 10 (K, 


SRGH). 

— 3228 (Butterworth): Idutywa (-AB), Reynolds 1041 (BOL). 

LESOTHO— 2828 (Bethlehem): Leribe (-CC), Dieterlen 245 (SAM, K). 

—2927 (Maseru): Mamathes (-BB), Guillarmod 585 (PRE); Mahlatsa, Berea, Gud- 
larmod 1595 (PRE); Molima Nthuse Pass (-BC), Schmitz 6230, 6283 (PRE); Rama 
campus, Schmitz 4171 (PRE). 

—2928 (Marakabei): Khatleleli (-BD), Dieterlen 245 (PRE). 

—2929 (Underberg): Sehlabathebe Nat. Park (-CC), Hoener 1858 (PRE)- . 

CAPE— 3126 (Queenstown): Queenstown distr., Sidey 3726 (S); Madeira, 


Queenstown (-DD), Thorns BOL 31740. 

—3127 (Lady Frere): Clarke’s Siding near Dordrecht (-AC), Bayliss 7942 (MO). 


Fig. 20 

Tritonia lineata var. lineata: a, plant and front view of flower; b, outer 
ner (right) bract; c, flower half; d, capsule. (Olivier ^603). 


(left) 


and in 


374 


Journal of South African Botany 


— 3225 (Somerset East): Klein Bruintjieshoogte (-CB), Drege 3499 (S); Bosberg 
(-DA), MacOwan 1199 (BOL, K, S). 

— 3226 (Fort Beaufort): Fairford (-BD), Cotterrell 50 (GRA); Alice (-DD), Gold- 
blatt 438, 370 (BOL). 

— 3227 (Stutterheim): Kingwilliamstown (-CD), Liebenberg 3141 (K), Tyson 3130 
(SAM); Komgha (-DB), Flanagan SAM 3612\ 10 mis from East London to Kingwil- 
liamstown (-DD), Marais 448 (BOL, GRA); East London, Rattray 192 (GRA), 
Sidey 1145 (S). 

— 3323 (Willowmore): 3-5 km W of Louterwater (-DC), De Vos 2567 (STE). 

— 3324 (Steytlerville): Near Kareedow (-CD), De Vos 2568 (STE); Baviaanskloof 
(-DA/-DB), Bayliss 6882 (MO). 

— 3325 (Port Elizabeth): Bronne Reservaat (-CD), Olivier 2603 (STE); Gravel road 
SW of Addo village (-DA), De Vos 2569 (STE); Port Elizabeth (-DC), Drege s.n., 
Oct. 1908 (GRA); Aerodrome, Long 155 (GRA, K); Baakens Valley, Olivier 833 
(NBG). 

—3326 (Grahamstown): Near Alicedale (-AC), Bayliss 6814, 7092 (MO); Cold- 
spring, Dlumane P62 (GRA); Fort Beaufort road, 10 mis from Grahamstown (-BA), 
Bayliss 2997 (MO, SRGH); Grahamstown (-BC), Daley 866 (GRA); Between 
Blaauw Krantz & Kaffir Drift (-BD), Burchell 3710 (K). 

— 3422 (Mossel Bay): Mossel Bay (-AA), Wall 47 (S). 

— 3424 Humansdorp): Seekoeirivier (-BB), Montgomery 1 (STE); Cape St Francis, 
Bayliss.n., 21/10/75 (MO). 

Without precise locality: Pappe s.n. (SAM); Drege 3499c (L); Irid. 99 (G, MO), 
101 (G), 280 (G); Thunberg 989 (UPS); Hb Burman sub Ixia folds linearibus (G); 
Burchell s.n. (K); Harvey s.n. (K); Kolbe BLFU 1187. 

Figure 9592 in Curtis’s Botanical Magazine (1939-40) is that of a yellow- 
flowered plant which is named Tritonia flavida. It was collected in Natal in 
1927 and cultivated for many years in England. It may be the yellow- 
flowered variation of var. lineata, although the stamens are not shown to be 
contiguous or curved towards the posterior side of the flower. 


b. var. parvifolia De Vos, var. nov. 

A varietate lineata foliis brevioribus et angustioribus, plerumque erectis, 
multo brevioribus quam scapo differt. 

Type: Cape: Jeffreys Bay, duineveld, Fourcade 3345 (STE, holo-; PRE, K, MO). 


Plants 150-200(-300) mm long. Scape 150-200(-300) mm long, usually 
unbranched, with l-2(-4) flowers. Leaves linear-lanceolate, usually erect, 
(40-)50-80(-150) x 2-5 mm, usually not reaching even near the base of the 
spike, drying to light brown. Style 15-18 mm. 

Distribution: south eastern Cape districts of Humansdorp, Uitenhage 
and Port Elizabeth. 

CAPE — 3325 (Port Elizabeth): Port Elizabeth, Holland BOL 31771; Addo Nat- 
ional Park (-DA), De Vos 2571 (STE); Zwartkops Rivier (-DB) Zeyher 3972 (SAM, 
K); On the hills of Adow and the fields near the Zwartkop River, Ecklon & Zeyher 
1021 (GRA partly, PRE, SAM, K), E. & Z. 101 (GRA, MO partly, not G). Perse- 


African genus Tritonia Ker-Gawler: part 2 


375 


verance (-DC), Rodin 1246 (BOL, K, MO); Inter Uitenhage & Bethelsdorp saltpan. 
Fries e.a. (SAM 68099 , K). 

— 3424 (Humansdorp): Jeffreys Bay and Ferreira Town near sea (-BB), Gillett 2319 
(STE); Jeffreys Bay duineveld, Fourcade 3345 (PRE, STE, K, MO); Humansdorp, 
Loubser 881 (NBG). 

Compared with the wide range of var. lineata, this variety has a restric- 
ted distribution in the south eastern Cape Province. It is distinguished by its 
small erect leaves and by flowers that are identical with the cream and straw- 
coloured variations of var. lineata and often have, like the latter, a small 
ridge-like callus on the median anticous perianth segment. It is not the same 
as var. minor which was described in Redoute’s Liliacees 7 sub t. 400. The 
latter has much longer leaves and is now incorporated with the typical var- 
iety. 

The leaves of var. parvifolia resemble those of T. strictifolia and poor 
herbarium specimens can easily be confused. The flowers, however, are dif- 
ferent and the two taxa have now been placed in separate sections of the 
genus. T. strictifolia, with its more zygomorphic flowers and high calli, is 
closely allied to T. securigera and T. laxifolia, and has therefore been placed 
in section Montbretia. 


14. Tritonia disticha (Klatt) Bak. Handb. Irid. 193 (1892) & in FI. Cap. 
6; 123 (1896). 

Plants (200-)500-800(-l 000) mm long. Corm depressed globose to 
ovoid, 10-30 mm diam.; tunic fibres elongated-reticulate, almost parallel to- 
wards the base. Scape simple or 1-3-branched, sometimes slightly flexuose, 
(200-)500-800(-1 000) mm long. Leaves 4-8, linear to linear-lanceolate, 
erect or suberect, acute or acuminate, (250-)350-500(-700) x 
(5_)8_12(-20) mm, with a strong middle vein and a strong vein near each 
edge, reaching, or sometimes not reaching, the base of the spike, usually 
drying to dark brown with pale veins; cauline leaves 2-4, 30-300 mm long. 
Spikes 1— 3(— 4), at first distichous, later subsecund, lax, with 4-12 or more 
flowers. Bracts membranous, golden-brown, darker brown upwards, with 
wide papery margins, (8-)10-15(-20) mm long; outer ovate or ovate-lanceo 
late, acute or acuminate or 3-toothed with the middle tooth sometimes re 
duced, or irregularly toothed, striate, with a strong dark median vein, the 
lower ones sometimes elongating after flowering; inner bidentate wit acu 
minate teeth, with 2 strong dark veins, often slightly shorter than the outer. 
Flowers slightly zygomorphic, funnel-shaped, (20— )25— 30(— 35) mm ong, 
bright red, orange-red or pink, with a small yellow, sometimes red-mar- 
gined, blotch on the anticous perianth segments, sometimes dark veil- 
ing. Perianth tube funnel-shaped, (8-)12-16 mm long, with the basal 3 mm 


376 


Journal of South African Botany 


narrow, widened upwards to 8-10 mm diam., subequal to or slightly longer, 
or rarely slightly shorter, than the segments; segments shortly elliptical 
or oblanceolate-obovate, obtuse, spreading, subequal in length, 
(10-)15-18(-20) mm long, the median posticous segment 8-10 mm wide, 
others 5-6 mm, the median anticous segment sometimes with a low yellow 
callus. Stamens curved towards the posticous side; filaments 7-12 mm long; 
anthers 5-7 mm long, curved, pink or yellow, reaching halfway or higher up 
the perianth segments. Style (15— )18— 22 mm; stigmatic branches 2-4 mm 
long, usually overtopping the anthers. Capsules subglobose or obovoid, ca. 
10 mm long; seeds brownish-black, shiny, minutely alveolate, 2 mm diam. 

Flowering period: (November to) December to February (to March). 

Distribution: from the south eastern Cape Province throughout Transkei 
and Natal to north eastern Orange Free State and south eastern Transvaal, 
and into Swaziland. (Fig. 16). 

This species, which now includes T. rubrolucens, is closely related to 
T. lineata, resembling it in general habit and especially in its long, erect or 
suberect, leaves which have a prominent vein near each edge and large 
amounts of tannin in the epidermis (De Vos, 1982a), usually causing the 
leaves, on drying out, to turn dark brown except for their pale veins. The 
distribution of the two species is almost similar. T. disticha differs from 
T. lineata in its red, pink or orange-red flowers with less distinct veining and 
with a yellow blotch in the throat on the anticous perianth segments, as well 
as in its different flowering period (summer). Faded herbarium specimens 
can hardly be distinguished when the flower colouring or flowering period is 
not mentioned. 

The species was first described by Klatt (1863) as T. rosea. This name is 
an illegitimate later homonym of T. rosea (Jacq.) Aiton (1810) which is 
T. flabellifolia. Foster (1936) renamed the species T. rubrolucens , being un- 
aware that Klatt had described Drege’s collection of the same species 
(Drege 4549) in B, as Tritonixia disticha. This species and T. rubrolucens are 
now treated as conspecific, as a wide range of intermediates which connect 
the two taxa occurs. 


Key to the Subspecies 

Outer bracts acute or acuminate, or obtuse and apiculate, usually not 3-toothed, 

the lower ones on the spike often elongating after flowering 

a. subsp. disticha 

Outer bracts 3-toothed, the middle tooth sometimes reduced, or irregularly 
toothed, or the lower ones sometimes acute, not elongating after flowering 
b. subsp. rubrolucens 


African genus Tritonia Ker-Gawler: part 2 


377 


a. subsp. disticha 

Tritonixia disticha Klatt in Abh. naturf. Ges. Halle 15: 356 (1882). Type: 
South Africa, Omsamwubo, Hohe am Fluss, Drege 4549 (B, holo-; P, S. 
iso-). See below. 

Tritonia disticha (Klatt) Bak. Handb. Irid. 193 (1892) & in FI. Cap. 6: 
123 (1896); Klatt in Dur. & Schinz, Consp. FI. Afr. 5: 205 (1895). 

T. coccinea L. Bol. in Ann. Bol. Herb. 4: 27 (1925); non Eckl. Top. 
Verz. 29 (1827). Type: Cape, Pondoland, Port St Johns, McNair s.n. NBG 
89/23 (BOL, holo-). 

Bracts up to 20 mm long; outer as described in key. Flowers (20-)30-35 
mm long, bright red or pink. Perianth tube subequal to or slightly longer 
than the segments. Style 14-20 mm long. 

NATAL-TRANSKEI — 3029 (Kokstad): Clydesdale (-BD), Schlechter 6620 (B, 
L. Z); Ingeli slopes (-DA), Strey 6378 (NH). 

NATAL — 3030 (Port Shepstone): Near Highflats (-AC), Grant 3568 (MO); Du- 
misa am Ifafas (-AD), Rudatis 228 (S); Melisa, Port Shepstone & Margate (-CB-D), 
Hafstrom & Acocks 333 (S); Port Shepstone (-CB), Sidey 3235 (S); Marburg, Thode 
5597 (STE); Izotsha Falls, Prosser 1418 (K); Paddock (-CC), Strey 6136 (NH); 
Mgongongo (-CD), Strey 9451 (NH). 

TRANSKEI— 3129 (Port St Johns): Omsamwubo, Hohe am Fluss (-AD/-DA), 
Drege 4549 (P, S); Port St Johns (-DA), Galpin 3443 (BOL), McNair NBG 89123 
(BOL), Moss 4317 (BM), Schonland 4094 (GRA); Lusikisiki (-BC), Eglin NBG 
2269/35 (BOL); Mt. Thesiger, (-CB), Hilliard 1072 (NU). 

—3130 (Port Edward): Port Edward, southern side (-AA), Bayliss 2516 (GRA, B, 
G), Strey 4506 (NH, PRE); Umtamvuna River, Thode 5099 (STE); 1 mile from Um- 
tamvuna River bridge, Codd 10702 (PRE). 

Without precise locality: Drege 4549 (B); Pondoland, Bachman 362 (BM); Wild 
coast, Transkei, Van der Riet NBG 30/63 (STE). 

Type: No precise locality is indicated on the holotype sheet, Drege 4549 
in B. On the isotype sheets in S and P, however, the locality is given as Om- 
samwubo. This is the river with its mouth at Port St Johns in the Transkei. 

This subspecies may be regarded as a low altitude form of the species, 
ranging from Port St Johns more or less to Margate in Natal, over a distance 
of some 120 kilometres. Several collections connect the two subspecies and 
some cannot be correctly placed, e.g. Hb Moss 16034 in BM and K, from 
Lusikisiki. See also under subspecies rubrolucens. 

Klatt (1882), followed by Baker (1896), incorrectly described the anthers 
as being twice as long as the filaments and the spike as distichous (whence 
the name). As in other species of Tritonia, the spike is at first, before the 
opening of the flowers, distichous, but later becomes subsecund. Baker also 
incorrectly described the perianth segments as being* twice as long a.-> 1 ie 


378 


Journal of South African Botany 


tube and used this character in his key to distinguish T. disticha from other 
species such as T. lineata. 


b. subsp. rubrolucens (Foster) De Vos, stat. nov. 

Tritonia rubrolucens Fost. in Contr. Gray Herb. Harv. N.S. 114: 146 
(1936), nom. nov. pro T. rosea Klatt, nom. illeg. Type as for T. rosea. 

T. securigera Ker-Gawl. sensu Eckl. Top. Verz. 29 (1827); non (Ait.) 
Ker-Gawl. (1805). 

T. rosea Klatt in Linnaea 32: 760 (1863) & in Dur. & Schinz, Consp. FI. 
Afr. 5: 207 (1895); Bak. Handb. Irid. 194 (1892) & in Curtis’s bot. Mag. 119 
t. 7280 (1893) & in FI. Cap. 6: 126 (1896); hom. illeg., non (Jacq.) Ait. 1810 
( = T. flabellifolia (Delaroche) Lewis). Type: Cape, Katberg, iiber den 
Waldern, Ecklon & Zeyher Irid. 98 (S, lecto-; SAM partly, G). See below. 

Montbretia rosea (Klatt) Bak. in J. Linn. Soc. 16: 169 (1877), non Voigt 
1845 (=T. flabellifolia). 

Tritonixia rosea (Klatt) Klatt in Abh. naturf. Ges. Halle 15: 356 (1882). 
leones: Curtis’s bot. Mag. 119 t. 7280; Batten & Bokelmann PI. 31/9; 
Pearse p. 83 Fig. 2. 


Bracts up to 15 mm long; outer as described in the key. Flowers 25-35 
mm long, bright red, orange-red or bright pink, sometimes with dark vein- 
ing. Perianth tube subequal to or slightly longer or shorter than the seg- 
ments. Style 18-22 mm long. 

Distribution: as indicated for the species. 

TRANSVAAL— 2730 (Vryheid): South Hill, Wakkerstroom (-AC), Galpin 9872 
(PRE, K); Oshoek, Wakkerstroom (-AD), Devenish 416 (NH, PRE), Mauve 4534 
(PRE); Kastrolnek, Van Dam 28 (PRE). 

ORANGE FREE STATE-NATAL— 2829 (Harrismith): Van Reenen (-AD), 
Wood 9521 (NH), Thode 3915 (STE); Harrismith, Manyenyezaberg, Jacobsz 514 
(PRE); Oliviershoek, Natal (-CA), Thode 9218 (STE); Cathedral Peak area (-CC), 
Schelpe 116 (NU), Robinson 56 (NU). 

NATAL — 2830 (Dundee): 51 km S of Nqutu on road to Kranskop (-DB), 
Davidse 6886 (S). 

—2831 (Nkandla): Ngoye (-DC), Huntley 720 (NU). 

— 2929 (Underberg): Giant’s Castle Game Reserve (-AD), Trauseld 584 (NU); Ver- 
gelegen Nature Reserve (-CB), Hilliard & Burtt 11185 (NU). 

— 2930 (Pietermaritzburg): Howick (-AC), Medley-Wood 9358 (BOL, PRE); Mt Gil- 
boa (-AD), Hilliard & Burtt 11867 (NU); Ahrens (-BB), Fisher 965 (NU, NH); 
Zwartkop (-CB), M. Wood 10240 (NH); Richmond (-CD), Saunders s.n. (K); Inan- 
da (-DB), M. Wood 384 (BM), 784 (NH), Johnson 1410 (NH); Bothas (-DC), M. 
Wood 12390 (PRE); Kloof (-DD), Lawson 680 (NU), Dohse e.a. 32 (NH); Hillcrest, 
Coleman 298 (NH). 

— 3029 (Kokstad): Insiswa (-CD), Schlechter 6471 (C, S, Z); Weza-Ngeli slopes 
(-DA), Strey 6378 (NU, PRE). 

— 3030 (Port Shepstone): Dumisa (-AD), Huntley 153 (NU); 10 mis inland from 


African genus Tritonia Ker-Gawler: part 2 


379 


Scottburg (-BC), Poynton 31 (NU); Oribi Flats (-CA), Rump s.n. (NU); Meongoeo 
(-CD), Strey 9451 (NH); Skyline, Uvongo, Nicholson NH 67241. 8 8 

—3130 (Port Edward): Port Edward (-AA), Huntley 722 (NU). 

TRANSKEI — 3128 (Umtata): Baziya (-CB), Baur431 (BOL, GRA, NBG, K) 
—3129 (Port St Johns): Fraser Falls (-BC), Leighton 2999 (BOL); Umkweni (-BD) 
Tyson 2623 (BOL, NBG). 


— 3228 (Butterworth): Kentani Valley (-CB), Pegler266 (BOL). 

SWAZILAND — 2631 (Mbabane): Hlatikulu (-CD), Stewart 57 (K) 2584 
(NBG). 

CAPE — 3226 (Fort Beaufort): 13.12, i.e. Winterberg zw. Tarka u. Katberg 
(-AD), Ecklon & Zeyher Irid 98 (SAM partly, G), Ecklon & Zeyher 13.12 (L, C); 
Winterberg, Zeyher s.n. (K); Katberg near Black Bridge (-BC), Dyer 2317 (GRA); 
Katberg, fiber den Waldern, Ecklon & Zeyher Irid. 98 (S); Katberg, Galpin 1704 
(GRA, PRE, K), Harvey s.n. (K); Katberg Pass, Adams 194 (NU); Hogsback, Alice 
(-DB), Barker 39 (BOL); Gaikaskop, Hogsback, Moss 15010 (BM), Young 988 (K). 
— 3227 (Stutterheim): Evelyn Valley, Leighton 2719 (BOL, GRA, PRE); Summit of 
mountain above Dohne (-CB), Flanagan 2306 (BOL, NBG); Cata Forest Res., Keis- 
kammahoek (-CA), Strey 3326 (PRE). 

— 3325 (Port Elizabeth): Port Elizabeth (-DC), Kemsley 1108/174 (GRA); Uitenhage 
pone Zwartkopsrivier, Pappe 239 (SAM, K). 

— 3326 (Grahamstown): 35 km voor Fort Beaufort vanaf Grahamstad komende 
(-BA), Von Teichman 369 (PRE); 10.10, i.e. Mundung Boesmansrivier (-DA), Eck- 
lon & Zeyher 99 (SAM). 

Without precise locality: Drege s.n. sub T. rosea (C); sub T. securigera (L); Brit- 
ish Kaffraria, Cooper 1807 (K); Orange Free State, Cooper 3201 (PRE, K); Said to 
be from Riversdale, Bayliss 2421 (G) (cultivated? wrong label?). 


Type: Klatt (1863) cited two syntypes: “Ex horto Hesse, leg. Bergius, 
leg. Krebs; leg. Ecklon & Zeyher, Irid. 98. Herb. Reg. Berol”. An answer 
from B to my enquiry stated that these syntypes are no longer there. The 
only isotypes found are Ecklon & Zeyher Irid. 98 in SAM, G and S. Those 
in G and S are the same species (in G from region 13.12, i.e. Winterberg 
zwischen Tarka und Katberg, and in S from 12.3, i.e. Katberg iiber den 
Waldern). The isotype in SAM comprises two species: two very poor speci- 
mens are similar to those in G and S, whereas the other may be T. lineata. 
As the specimens in SAM are so poor, Ecklon & Zeyher Irid 98 in S is 
chosen as lectotype. 

This subspecies is very closely related to, and more or less sympatric with 
T. lineata, differing in its later flowering period (summer), in flower colour- 
ing and a less pronounced veining on the perianth, and often in a slightly 
longer perianth tube. Flower colouring varies from pink to bright red or 
orangy-red. Baur 431 in BOL from Baziya in the Transkei has pinkish flow- 
ers. Schelpe 5069 in BOL, from near Port Edward, has “yellow or light red" 
flowers. Whether this collection constitutes an unusual colour rm or hy- 
brids between T. lineata and subspecies rubrolucens, is uncertain. Its flower- 
ing period is that of subspecies rubrolucens. 


380 


Journal of South African Botany 


The outer bracts of this subspecies vary. In some collections, e.g. Moss 
15010 in BM, from Gaika’s Kop near Alice, the outer bracts of the lower- 
most flowers of the spike are acute or acuminate and approximate bracts of 
subspecies disticha. 

5. Section MONTBRETI A 

Sect. Montbretia (DC.) Pax in Nat. Pfl. Fam. 2 (5): 155 (1888) in part, 
excl. T. rosea; Diels, ibid. ed. 2, 15a: 491 (1931). 

Montbretia DC. in Bull. Sci. Soc. Philom. 3: 151 (1803); Reichenb. 
Consp. Regni Veg. 1: 60 (1828); Klatt in Linnaea 32: 752 (1863) & in Abh. 
naturf. Ges. Halle 15: 359 (1882); Bak. in J. Linn. Soc. 16: 167 (1877) in 
part, excl. all spp. except M. securigera, M. laxifolia, M. strictifolia; Stapf in 
Curtis’s bot. Mag. 150 sub t. 9038 (1924). 

Tritonia subgenus Montbretia (DC.) Bak. Handb. Irid. 190 (1892) & in 
FI. Cap. 6: 199 (1896). 

Type species: T. securigera (Ait.) Ker-Gawl. 

Leaves lanceolate to linear, rarely with undulate margins, rarely slightly 
succulent. Inflorescence a simple or few-branched spike or rarely reduced to 
a few terminal stalked flowers (T. florentiae). Flowers zygomorphic, some- 
what 2-lipped or sometimes funnel-shaped or rarely almost salver-shaped, 
red to orange, salmon or yellow, sometimes pink or cream. Perianth tube 
funnel-shaped or tubular, slightly shorter to up to four times as long as the 
segments; segments oblong or oblanceolate, often subequal in length, the 
posticous segment usually widest, often concave and erect, the three anti- 
cous segments, or rarely only the median anticous segment, with a large yel- 
low callus in or near the throat. Stamens and style curved towards or against 
the posticous segment; anthers usually contiguous, facing the anticous side 
of the flower. 

Distribution: widespread in the Transvaal, southern and eastern Cape 
Province including the Transkei, and Great Karoo; also in Namaqualand as 
well as in Mozambique (T. moggii ), Malawi, Zambia and Tanzania (T. laxi- 
folia); excluded from the south western Cape and Natal (Figs 21, 22). 

The essential characters of this section are the presence of a large yellow 
callus on each of the anticous perianth segments, or rarely on the median 
segment only, and a decided zygomorphy with the posticous perianth seg- 
ment often widest and concave, thus protecting the stamens and stigmas. 

The presence of calli in two species, at that time still under Gladiolus , 
was the reason that A.P. De Candolle (1803) established a new genus, 
Montbretia (the name to commemorate Coquebert-Montbret the younger). 
This genus was at first recognised by Baker (1877) and Klatt (1863, 1882); 


African genus Tritonia Ker-Gawler: part 2 


381 


but later, after Bentham had assigned it to the synonymy of Tritonia in Ben- 
tham and Hooker’s Genera (1883), Baker (1892, 1896) and Klatt (1895) re- 
garded it as a section or subgenus of Tritonia. In the twentieth century Stapf 
(1924) and Sealy (1939-40) proposed that Montbretia be re-established as 
genus. Phillips (1951) on the other hand, as well as Goldblatt (1976), did not 
agree with this, and during the present investigation numerous intermediates 
with small calli or ridges on the anticous perianth segments were found 
amongst other sections of Tritonia-, therefore Montbretia is here also includ- 
ed with Tritonia as a section. 

The calli which are often situated on small yellow, red-margined 
blotches, vary in shape in different species (De Vos, 1982b), and were it not 
that they are often not readily visible in herbarium material, a better use 
could be made of this character to distinguish certain species. The only func- 
tion that can possibly be attributed to the calli is that they narrow the throat 
of the perianth, thus ensuring that pollinating insects will brush with their 
backs against the anthers and stigmas. There is no evidence that the calli are 
nectaries. These are in the septa in the top of the ovary. 

Several species, such as T. watermeyeri, T. securigera, T. chrysantha, 
from the southern and eastern limits of the winter rainfall region, as well as 
two species from the northern limits of this region (Richtersveld in Nama- 
qualand) are winter rainfall species, sprouting in autumn and flowering in 
spring and early summer. Another two species, namely T. florentiae and T. 
karooica, from the Great Karoo, have retained their winter growing habit 
and also flower in spring. The remaining five species are from the summer 
rainfall regions of the eastern Cape and Transvaal, and countries farther 
north. They sprout in spring or summer. Of these T. nelsonii, T. atrorubens 
and T. drakensbergensis flower in summer, T. moggii from Mozambique in 
autumn or winter, and T. laxifolia in autumn. The last named species has a 
disjunctive distribution: in the eastern Cape Province and in Central Africa, 
with apparently a wide interval between these areas. 

This section can be subdivided into several groups or subsections. Most 
of the species are very closely related and have almost similar flowers: some- 
what two-lipped with a large posticous perianth segment against which the 
stamens and style are curved, and large anticous calli. Nine species can be 
placed in this group, e.g. T. securigera, T. laxifolia, T. nelsonii, T. water- 
meyeri, etc. They are from the southern and eastern Cape, Transvaal and 
farther north. The two species from the Richtersveld, namely T. marlothii 
and T. delpierrei, have somewhat salver-shaped flowers and slightly succu- 
lent leaves and form a separate group. The two species from he Great 
Karoo are very different from each other: T. florentiae has an unusual inflo 
rescence; and T. karooica has funnel-shaped flowers. Each could be as 
signed to separate subsections. 


382 


Journal of South African Botany 



A T.strictifolia 
• T. securiga 
■ T. watermeyeri 



Fig. 21 

Geographical distribution of Tritonia , section Montbretia : T. chrysantha, T. floren- 
tiae, T. parvula, T. securigera, T. strictifolia, T. watermeyeri. 


Formal subsections are, however, not described, as T. atrorubens cannot 
be placed, on account of somewhat inadequate herbarium material. 

Section Montbretia is not the Montbretia of horticulturalists which con- 
sists of species of Crocosmia and (or) Crocosmia hybrids. 


African genus Tritonia Ker-Gawler: part 2 


383 




Fig. 22 

Geographical distribution of Tritonia, section Montbretia: T. atrorubens, T delpierrei, 
T. drakensbergensis, T. karooica, T. laxifolia, T. marlothii, T. moggii, T. nelsonii. 


384 


Journal of South African Botany 


15. Tritonia securigera (Ait.) Ker-Gawl. in Kon. & Sims, Ann. Bot. 1: 
228 (1805); Bak. Handb. Irid. 196 (1892) & in FI. Cap. 6: 128 (1896); Klatt 
in Dur. & Schinz, Consp. FI. Afr. 5: 207 (1895). 

Gladiolus securiger Ait. Hort. Kew. 1: 65 (1789); Curtis’s bot. Mag. 11 
t. 383 (1797). Type: ex Hort. Chelsea, 1778, 5. loc., s. coll. (BM, lecto-). 
See below. 

Montbretia securigera (Ait.) DC. in Bull. Soc. Philom. 3: 151 (1803) & in 
Red. Lil. t. 53 (1804); Klatt in Linnaea 32: 752 (1863) & in Abh. naturf. 
Ges. Halle 15: 359 (1882); Bak. in J. Linn. Soc. 16: 169 (1877). 

Ixia gladiolaris Lam. Encycl. 3: 341 (1789). Type: Hb. Lam. (P). 

I. squalida Thunb. in Hoffm. Phyt. Blatt. 4 (1803) & FI. Cap. 1: 244 
(1811) horn, illeg., non Ait. Type: HB. Thunberg 996 (UPS). 

I. flabellularis Vahl, Enum. PI. 2: 67 (1805). Type: Cap. b. Spei, ex herb. 
Royeni (C, holo-). 

I. thunbergii Roem. & Schult. Syst. Veg. 1: 391 (1817), nom. nov. pro 
I. squalida Thunb. Type as for I. squalida Thunb. 

Tritonia flava (Ait.) Ker-Gawl. sensu Ker-Gawl. in Bot. Reg. 9 t. 747 
(1823); Bak in FI. Cap. 6: 128 (1896). (Basionym Gladiolus flavus Ait. =Tri- 
tonia sp. indet. — see Species Insuffieciently Known p. 412) 

leones: Curtis’s bot. Mag. 11 t. 383; Red. Lil. t. 53; an unpublished 
figure by M. Page in BOL; this work Fig. 23/2. 

Type: The type material in BM comprises two sheets, one with two 
plants, e Hort. Chelsea 1778, without collector or locality, and the other 
with a single flowering scape, from Cap. b. Spei, without collector or date. 
These are pobably plants or descendants of plants, brought back from the 
Cape by Masson. The first is chosen as lectotype. 

Plants (100-)150-300(-350) mm long. Corm subglobose, 10-20 mm 
diam.; tunic fibres fine or sometimes rather strong, matted, reticulate in the 
upper part, almost parallel in the lower. Scape (100-)150-300(-450) mm 
long, erect or suberect, simple or with one branch, sometimes with a collar 
of old leaf bases around the base. Leaves 4-7 , suberect or rarely spreading, 
lanceolate or lanceolate-linear, (70-)100-200(-270) x (4 — )6— 10(— 12) mm, 
acute or acuminate, obscurely veined, with a slightly stronger middle vein, 
not reaching the spike or sometimes as long as the spike; cauline leaves 1-2, 


Fig. 23 

Tritonia laxifolia (De Vos 2443): la, plant and front view of flower; lb, dehiscing 

capsule. 

Tritonia securigera (De Vos 2417): 2a, plant and front view of a flower; 2b, outer 
(left) and inner (right) bract; 2c, flower half in a median longitudinal cut; 2d, dehisc- 
ing capsule. 


African genus Tritonia Ker-Gawler: part 2 


385 





386 Journal of South African Botany 

up to ca. 60 mm long. Spike erect, rather lax, almost secund, with 5-15 
flowers. Bracts membranous, papery, ovate, 7-10(-12) mm long, pale with 
dark veins, brownish towards the top; outer 3- or 5-toothed or irregularly 
toothed, veined; inner 2-dentate, 2-veined, sometimes slightly shorter than 
the outer. Flowers suberect and slightly spreading, (28-)30-35(-45) mm 
long, reddish-orange or apricot, with a yellow throat, or almost wholly yel- 
low (RHS 28B, C, 31B, 26B, 23C, 19B, 21D), with the upper lobe some- 
times pale yellow inside. Perianth tube tubular in the lower half, funnel- 
shaped in the upper, (10-)12-15(-20) mm long, slightly curved; segments 
unequal, obtuse, 12-20 mm long, the posticous segment obovate, often 
slightly concave, (8-) 12- 15 mm wide; other segments oblong to oblanceo- 
late, spreading, upper laterals 7-11 mm wide, anticous segments 5-8 mm 
wide, each with a high, and usually wide, callus in the yellow red-margined 
throat. Stamens curved towards the posticous segment; filaments 9— 12(— 15) 
mm; anthers contiguous 5-6 mm long, purple or mauve, reaching halfway or 
sometimes higher up the perianth lobes; pollen mauve. Style slightly bent, 
17-20(-25) mm long; stigmatic branches 4-5 mm long, often overtopping 
the anthers. Capsules ellipsoid, 7-10 mm long; seeds brown, shiny, angled, 
reticulate-foveate, 1,5-2 mm diam. 

Flowering period: September to early December. 

Distribution: in the Little Karoo, Langkloof, eastern Karoo and south- 
ern Cape coastal districts of Mossel Bay, George and Humansdorp (Fig. 
21 ). 

CAPE — 3224 (Graaff-Reinet): Kamdebo (-AB), Dunn BOL 31758. 

— 3225 (Somerset East): Prope Middleton (-DD), Rogers BOL 12945. 

— 3226 (Fort Beaufort): 17 km E of Adelaide (-CD), De Vos 2573 (STE). 

— 3321 (Ladismith): 5-10 mis W of Ladismith (-AD), Ry croft 3029 (NBG); Foothills 
of Langeberg N of Herbertsdale (-DD), Goldblatt 4155 (MO). 

—3322 (Oudtshoorn): Kleine Poort, Tyson 3071 (PRE, SAM, K); Eikerkraal betw. 
Klaarstroom & Prince Albert (-AB), Leipoldt BOL 31763; Foot of Swartberg Pass 
on road to Oudtshoorn (-AC), Goldblatt 2958 (MO); Between Cango and DeRust 
(-AD), Thorne SAM 53367; Oudtshoorn (-CA), Yeoman s.n., 17/5/25 (NBG), Deas 
SAM 11639 ; Ruiterbos (-CC), Barker 7677 (NBG); Hills near Zebra (-CD), Taylor 
850 (NBG); Camfer, De Vos 2416 (STE); George, Marloth 2455 (PRE); Near 
George, Burchell 6090 (K), Compton 13214 (NBG), Goldblatt BOL 31762 ; Near 
Blanco, Middlemost 1785 (BOL, NBG, STE, S); Between Grootbraak River & 
George, Fourcade 4074 (STE); George forest. Barker 6858 (NBG); Eseljagsrivier 
(-DC), De Vos 2417 (STE); Eseljagspoort, S end, Oliver 3584 (PRE); Betw. Eseljags 
& Molenrivier, De Vos 2418 (STE); 56 km E of Oudtshoorn, Goldblatt 2855 (MO); 
Oudtshoorn to Avontuur, Goldblatt 450 (BOL). 

— 3323 (Willowmore): Modderfontein, distr. Willowmore, Andreae 1042 (PRE); 
Zuurberg at Georgida (-AD), Fourcade 4629 (STE); Banks of Wagenboomsrivier 
near Joubertina (-DD), Fourcade 2331 (K). 

— 3325 (Port Elizabeth): Kommadagga (-BB), Bayliss 6085 (PRE, MO, S). 

— 3421 (Riversdale): Riversdale (-AB), Bayliss 2421 (G, not in SAM). 

— 3422 (Mossel Bay): Brak River (-AA), Hops 250 (BOL); Great Brak, slopes 
above station, Lewis 5607 (NBG, STE); Near Mossel Bay, Bowie 1 (BM). 


African genus Tritonia Ker-Gawler: part 2 


387 


Without precise locality: Thunberg 996 (UPS); Ecklon ex Mont. tab. pr. b. Sp 
(L) (probably a wrong locality label); Herb. Royeni (C). 

This is one of the first of the callus-bearing species of Tritonia found in 
South Africa, for which Aiton (1789) coined the appropriate epithet of secu- 
regera (axe-bearing). 

The species is closely related to T. laxifolia from which it can best be dis- 
tinguised by its different flowering period. Other points of difference, often 
not evident in herbarium material, are the yellow throat of the perianth, 
usually broader and less high calli, shorter stamens which usually do not 
reach more than halfway up the perianth segments, and shorter capsules. 
The two species occupy neighbouring areas in the south eastern Cape Prov- 
ince; towards the eastern extremity of its range T. securigera often shows 
calli that are higher than usual, approximating those of T. laxifolia. 

T. securigera is also closely related to T. chrysantha and T. parvula. It 
differs from the latter in its usually wider, firmer leaves, somewhat larger 
flowers with slightly longer perianth tube, stamens and style, and with the 
posticous perianth segment not, or only slightly, concave. It differs from 
T. chrysantha in its shorter perianth tube and usually in flower colouring 
which varies from a deep reddish-orange or apricot to salmon or various 
shades of yellow. The yellow-flowered form of T. securigera can be distin- 
guished from T. chrysantha only by its shorter perianth tube. 

The figure identified as T. flava in Botanical Register 9 t. 747 probably 
represents the yellow-flowered variation of T. securigera. This figure, being 
that of a plant from a later introduction into Britain, does not correspond 
with the holotype of T. flava (Ait.) Ker-Gawler in BM which, on account of 
its poor condition, cannot be connected with any species of Tritonia. See 
also under Species Insufficiently Known. 

A wide variation in size occurs in T. securigera. Some collections, e.g. 
Fourcade 4094 and 5054 from the Great Brak and George vicinity, consist of 
small plants 100-150 mm long with short, spreading leaves and few flowers; 
whereas specimens from the eastern Karoo, widespread between Fort Beau- 
fort and Pearston, e.g. De Vos 2573, as well as Bayliss 6085 from Komma- 
dagga, are up to 350 mm long with bracts to 15 mm and flowers 30-40 mm, 
with a perianth tube of 15-20 mm, and long leaves with a slightly stronger 
vein near each edge. As smaller plants were observed amongst the large 
plants, the latter have not now been described as a distinct variety. 

16. Tritonia chrysantha Fourcade in Trans. Roy. Soc. S. Afr. 21: 77 (1932). 
Type: Cape, Uniondale, hills S of Georgida, Fourcade 4394 (BOL, holo-; 

STE, K). 

Fig. 24. 




388 


Journal of South African Botany 


Plants (100-)200-300(-500) mm long. Corm subglobose to ovoid, 10-20 
mm diam.; tunic fibres strong, reticulate, almost parallel and plate-like to- 
wards the base. Scape (100-)200-300(-500) mm long, erect or slightly bent, 
simple or with 1-2 branches, often with dried leaf bases 40-100 mm long at 
the base. Leaves 4-7, lanceolate, acute, 50-100(-200) x (2-)5-10 mm, sub- 
erect or sometimes spreading, usually not reaching the base of the spike; 
cauline leaves 1-2, subulate. Spike distichous, lax, sometimes with 1-2 
branches, with 2-8 flowers. Bracts membranous, 6-12 mm long; outer 3- or 
2-toothed, with obtuse or acute teeth, with reddish-brown veins; inner 2- 
toothed, acute or acuminate, 2-veined, sometimes slightly shorter than the 
outer. Flowers (30-)35-40(-48) mm long, bright yellow (RHS 8B). Perianth 



Tritonia chrysantha (De Vos 2419): a, plant and front view of flower; b, outer (left) 
and inner (right) bract; c, flower half in a median cut; d, capsules. 


African genus Tritonia Ker-Gawler: part 2 


389 


tube (18-)20-28 mm long, tubular, widening gradually to 6-7 mm diam. in 
the upper quarter; segments ovate-spathulate, obtuse, (8— )10— 15 mm long, 
the posticous segment 9-12 mm wide, often slightly hooded, the others 5-8 
mm wide, spreading, the three lower ones with a large yellow callus up to 5 
x 3 mm. Stamens curved towards the posticous segment; filaments 7-10 
mm; anthers 4-6 mm long, violet or yellow, reaching halfway or higher up 
the perianth lobes. Style 20-30(-35) mm; stigmatic branches 3-5 mm long, 
sometimes overtopping the anthers. Capsules ca. 6 mm long; seeds shiny 
brown, 2 mm diam. Chromosome number 2n = 22. 

Flowering period: late August to October. 

Distribution: chiefly in the Little Karoo and southern and south eastern 
Great Karoo (Fig. 21). 

CAPE — 3225 (Somerset East): Zuurberg mts. near Somerset East, Bayliss 7132 
(MO). 

— 3322 (Oudtshoorn): Slopes east of Laudina (-DB), Oliver 3588 (STE, PRE); Be- 
tween Buffelsdrif & Bassau (-DB), De Vos 2419 (STE). 

— 3323 (Willowmore): Roodeklip SW of Anthoniesberg (-AD), Taylor 868 (NBG); S 
of Georgida, Fourcade 4394 (STE, BOL, K); Farm Vaalwater SE of Willowmore 
(-BC), Thompson 2008 (STE, K); Near Uniondale (-CA), Zinn SAM 65867, Yates 
PRE 36625; Hoogtes Pass, Potjiesrivier, Acocks 21593 (PRE). 

— 3324 (Steytlerville): Baviaanskloof (-CA), Bayliss 4954 (NBG, MO). 

— 3326 (Grahamstown): Near Alicedale (-AC), Bayliss 8047 (STE, MO), Bayliss 
6811 (MO); Trappes Valley (-BD), Bayliss 2404 (MO). 

— 3421 (Riversdale): The Fisheries (-BD ), Acocks 21565 (NBG, PRE). 

This species is distinguished by bright yellow flowers with a narrow peri- 
anth tube about twice the length of the segments and a large yellow callus 
often on a small pinkish-red blotch, on each of the anticous perianth seg- 
ments. It is closely related to T. securigera and can be confused with the yel- 
low-flowered form of this species, from which it differs in its longer perianth 
tube. 


17. Tritonia parvula N.E.Br. in Kew Bull. 1931: 452. Type: Cape, near 
Oudtshoorn, Muir 4724 (K, holo-). 

Plants (80-)150-250(-300) mm long. Corm 5-10 mm diam., ovoid; tunic 
fibres strong, reticulate in the upper half, almost parallel in the lower. Scape 
(60-) 120-200(-250) mm, slender, curved or flexuose, rarely branched. 
Leaves 4-7, linear, acuminate, somewhat flaccid, suberect or spreading, 
sometimes recurved, 80-300 x (0,5-)l-2(-4) mm, often longer than the 
scape, with a slender middle vein. Spike lax, often flexuose, w th (1 ) 
flowers, rarely with one branch. Bracts membranous, 5—9 mm long, pa e 
brown, with darker veins; outer minutely 3-toothed, darker brown upwar s, 


390 


Journal of South African Botany 


inner minutely bidentate. Flowers zygomorphic, (18-)22-30 mm long, red- 
dish-orange to salmon, with a yellow throat. Perianth tube tubular in the 
lower half, funnel-shaped above, 6— 10(— 13) mm long, widened to ca. 6 mm 
diam. at the throat; segments oblanceolate, obtuse, 9— 12(— 15) mm long, the 
posticous segment concave and hooded, often paler inside 6— 10(— 12) mm 
wide, the other 4-6 mm long, with a high yellow callus on each anticous seg- 
ment. Stamens curved towards the posticous segment; filaments 6-10 mm; 
anthers 4-5 mm long, curved, often violet, reaching below to above the mid- 
dle of the perianth segments. Style (10-)15-20 mm; stigmatic branches 2,5-3 
mm long, reaching or overtopping the anthers. Capsules shortly ellipsoid, 
ca. 6 mm long. 

Flowering period: late September to October (to December). 
Distribution: southern districts of the Cape Province from Riversdale to 
Willowmore Fig. 21). 

CAPE — 3321 (Ladismith): Seweweekspoort (-AD), Compton 7473 (NBG); 
Wurts 1607 (NBG); Klein Swartberg near Hoekoe, Wurts 1476 (NBG). 

— 3322 (Oudtshoorn): Oudtshoorn (-CA), Barker 7604 (NBG), SAM 53185, BOL 
31776, Muir 4724 (K). 

— 3322 Willowmore): Boskloof, Baviaanskloof (-DA), Hugo 1444 (STE); 2,2 mis 
from Louterwater to Oudtshoorn (-DC), Marsh 1406 (PRE, STE). 

— 3421 (Riversdale): Rock faces at Glen (-AB), Muir BOL 31773. 

Without precise locality: Oudtshoorn flower show, SAM 9392 partly. 

T. parvula is very closely related to, and largely sympatric with, T. secu- 
rigera. It differs from the latter in its more flaccid appearance, its narrow, 
flaccid, linear leaves, a somewhat flexuose spike that has slightly smaller 
flowers with smaller floral parts and with the median posticous perianth seg- 
ment strongly concave and the call! on the lateral anticous segments some- 
times reduced in size. Hugo’s collection no. 1444 in STE, from Baviaans- 
kloof, seems to connect the two species. Here the spike and flowers 
approximate those of T. securigera , but the leaves are only 2 mm wide. 
T. parviflora might perhaps be regarded as a depauperated variety of 
T. securigera. 

18. Tritonia strictifolia (Klatt) Benth. ex Klatt in Dur. & Schinz, Consp. FI. 
Afr. 5: 208 (1895). 

Montbretia strictifolia Klatt in Linnaea 32: 753 (1863) & in Abh. naturf. 
Ges. Halle 15: 359 (1882); Bak. in J. Linn. Soc. 16: 169 (1877). Syntypes: 
Cape, Albany, auf grasreichen Hiigeln bei Grahamstown, Ecklon & Zeyher 
100 (B, lecto-; SAM, G, S, Z); Uitenhage, Kalkhohe zw. Zwartkoprivier u. 
Zondagsrivier, Ecklon & Zeyher 260 (B, not seen). 

Tritonia laxifolia Benth. ex Bak. var. strictifolia (Klatt) Bak. Handb. 
Irid. 195 (1892) & in FI. Cap. 6: 127 (1896). 


African genus Tritonia Ker-Gawler: part 2 


391 


T. parviflora Bak. in Bull. Herb. Boiss. Ser. 2, 1: 865 (1901). Type: Kap- 
kolonie, Ecklon & Zeyher 100 (Z, holo-; SAM, B, G, S). 

Plants 150-200(-270) mm long. Corm 10-20 mm diam.; tunic fibres fine, 
almost parallel in the lower half, reticulate in the upper. Scape 150-200 
(-270) mm long, slightly flexuose towards the top, often with one branch. 
Leaves 4-7, linear-lanceolate, usually erect, (40-)60-90 x 2-5 mm, acute or 
acuminate, drying to a light brown, with a prominent vein near the margins 
and in the middle, not reaching near the spike; cauline leaf one, ca. 15 mm 
long. Spike rather lax, often slightly flexuose, somewhat distichous, with 3-5 
flowers, often with a 1-2-flowered branch. Bracts membranous, brown, 
7-10 mm long; outer striate, 3-toothed; inner bidentate, with acuminate 
teeth. Flowers zygomorphic, (20-)25-35 mm long, rust-orange or pink. 
Perianth tube 10-15 mm long, funnel-shaped, widened to ca. 8 mm diam. 
upwards; segments obtuse, 13-18 x 5-10 mm, with the posticous segment 
widest, anticous segments with a high, narrow, yellow callus on a yellow 
mark. Stamens curved towards the posticous segment; filaments 7-10 mm; 
anthers contiguous, ca. 5 mm long, reaching halfway or higher up the peri- 
anth segments. Style 17-20 mm; stigmatic branches 3-4 mm long, reaching 
or overtopping the anthers. 

Flowering period: November. 

Distribution: south eastern Cape Province, Grahamstown district (Fig. 

21 ). 


CAPE— 3326 (Grahamstown): Auf grasreichen Hiigeln bei Grahamstown (-BC), 
Ecklon & Zeyher 100 (SAM), Irid. 100 (B, G, S, Z); Between Grahamstown and 
Botha’s Hill, Hutchinson 1586 (BOL, K); Commonage N of Grahamstown, Barker 
6940 (STE); Grahamstown flats, Galpin 386 (PRE); 4 mis NE of Grahamstown, Gil- 
lett2517 (BOL, STE). , , „ . r . 

Without precise locality: Ecklon s.n. sub T. squalida (C); Ex hb. Royen sub Ixia 
(Morphixia) capillaris B stricta (C). 

This species comprises small plants with narrow, erect, linear-lanceolate 
leaves often less than 100 mm in length and usually less than 5 mm wide, 
and with few, short-tubed flowers resembling those of T. securigera and 
T. laxifolia. The leaves have three prominent veins, two of which are near 
the margins. The calli on the anticous perianth segments are high and nar- 
row as in T. laxifolia and the eastern form of T. securigera. The two latera 
calli are, however, sometimes reduced in size. Variation occurs in flower 
colouring which was described as red (“miniata ) by Klatt (1863), ut ater 

collectors noted the colour as rust-orange or pink. 

The species is related to T. laxifolia and T. securigera, i ering in i s 
smaller, erect leaves with their strong submarginal veins, and from the tor- 


392 


Journal of South African Botany 


mer also in its different flowering period. As the leaves are so distinctly dif- 
ferent and the flowering period different from that of T. laxifolia with which 
Baker (1896) associated it as a variety, T. strictifolia is once again raised to 
specific rank. 

In habit T. strictifolia resembles T. lineata var. parvifolia and herbarium 
specimens can be confused, nothwithstanding the difference in flower struc- 
ture. This variety occurs in the districts of Port Elizabeth and Uitenhage 
which adjoin the Grahamstown district. Klatt’s second syntype, Ecklon & 
Zeyher 260, from Kalkhohe in the Uitenhage district, could not be found 
during the present investigation. There is some slight possibility that it may 
be this small variety of T. lineata , as its locality, as well as its seeming pref- 
erence for chalky ground, is that of the latter. 

19. Tritonia laxifolia (Klatt) Benth. ex Bak. Handb. Irid. 195 (1892) & in FI. 
Cap. 6: 126 (1896) excl. var. strictifolia ; Klatt in Dur. & Schinz, Consp. FI. 
Afr. 5: 206 (1895); N.E.Br. in J. Linn. Soc. 48: 47 (1928). 

Montbretia laxifolia Klatt in Linnaea 32: 754 (1863) & in Abh. naturf. 
Ges. Halle 15: 359 (1882); Bak. in Trans. Linn. Soc. 29: 155 & t. 101A 
(1875) & in J. Linn. Soc. 16: 169 (1877); Stapf in Curtis’s bot. Mag. 150 t. 
9038 (1928) as M. laxiflora (sic), excl. syn. M. strictifolia. Syntypes: Cape, 
C.B.S., Krebs s.n.; Drege 3489a (K, lecto-; SAM, G, L, MO, S); Locality 
103, i.e. Uitenhage, Steenbokvlakte, Ecklon & Zeyher Irid. 105 (B, G, 
SAM, S, MO). 

Ixia fenestrata Jacq. sensu Thunb. in Hoffm. Phyt. Blatt. 4 (1803) & in 
FI. Cap. 244 (1811); non Jacq. (1789 & 1809). 

Tritonia bakeri Klatt in Dur. & Schinz, Consp. FI. Afr. 5: 203 (1895); 
non Klatt (1882). Type: German East Afr. 5° 50' S, 30° E, Speke & Grant 
s.n. (K). 

T. clusiana A. Worsley in Gard. Chron. 38: 269 (1905). Type: South 
Africa, said to have come from the Greytown distr., Worsley s.n. (K, holo-). 

T. bracteata A. Worsley in Gard. Chron 39: 2 (1906). Type as for T. clu- 
siana. 

leones: Curtis’s bot. Mag. 150 t. 9039; an unpublished drawing by G. J. 
Lewis (BOL); Trans. Linn. Soc. 29 t. 101A (incorrect, calli not shown); this 
work Fig. 23/1. 

Plants (150-)200-400(-600) mm long. Corm subglobose, 10-15 mm 
diam.; tunic fibres very fine, reticulate in the upper part, subparallel towards 
the base, often with small basal cormlets. Scape 140-400(-600) mm long, 
erect or suberect, simple or with 1-2 or rarely more branches. Leaves 4-7 , 
lanceolate or linear-lanceolate, acute or acuminate, suberect or sometimes 
spreading, 100-300(-500) x 5-10(-12) mm, with a strong middle vein and 


African genus Tritonia Ker-Gawler: part 2 


393 


sometimes a slightly stronger lateral vein near each edge, often reaching the 
base of the spike; cauline leaves 1-2, shorter and narrower. Spike somewhat 
lax, secund, sometimes 1-2-branched, with 6-12 or more flowers. Bracts 
membranous, reddish-brown, often with fine brown stippling towards the 
top, with wide, brownish, papery margins, 6— 12(— 18) mm long; outer 3- 
toothed or rarely 5-toothed or subacute, striate, with a stronger median vein 
ending in a dark, sometimes reduced and short middle tooth; inner biden- 
tate, the teeth with dark brown tips, 2-veined, sometimes slightly shorter 
than the outer. Flowers suberect and slightly spreading, (22-)25-32(-35) 
mm long, bright salmon-pink to pale brick-red (RHS 33C, 34C, D) the cup 
and sometimes the posticous perianth segment pale inside. Perianth tube 
narrowly tubular in the lower half, widely funnel-shaped in the upper, slight- 
ly curved, 12-16 mm long; segments unequal in width, obtuse, (10— )12— 15 
mm long, median posticous segment obovate to subrotund, 8-12 mm wide, 
suberect, concave and slightly hooded; other segments elliptical, 5-8 mm 
wide, the three anticous segments with a narrow, curved, yellow or green- 
ish-yellow, callus 3-5 mm high, on a small yellow, red-margined blotch. Sta- 
mens curved against the posticous perianth segment; filaments 12-15 mm 
long; anthers 4-6 mm long, contiguous, reaching to the upper quarter of the 
posticous segment, mauve or purple; pollen pale. Style 18-20(-22) mm long; 
stigmatic branches 2-4 mm long, reaching the anther tips or just above or 
below them. Capsules ellipsoid and somewhat trigonous, erect, up to 15 mm 
long; seeds brown, finely ribbed, ca. 2 mm diam. 

Flowering period: (January to) March to May (to July, Central Africa). 

Distribution: eastern Cape Province from Uitenhage to East London and 
eastern and north eastern parts of the Great Karoo, as well as in the Central 
African states of Malawi, Zambia and Tanzania (Fig. 22). 

CAPE— 3126 (Queenstown): Mountain side, Queenstown (-DD), Galpin 2036 
(GRA, PRE, K); Junction farm, Queenstown, Galpin 8187 (PRE). 

— 3224 (Graaff-Reinet): Along Sunday River N of Monkey Ford (-BA), Burchell 
2867 (K, L); Prope Graaff-Reinet (-BC), Bolus 597 (BOL, K), Bayliss 4457 (MO); 
Ad pedes Spandouw Kop, Francis 17 (BOL). 

—3225 (Somerset East): Cradock (-BA), Brincker TM 20243 (PRE), Bot. Hon. 306 
(MO); Pearson (-CA), Palmer s.n., Apr. 1962 (PRE); Ad pedis montis Boschberg 

(-DA), MacOwan 275 (K). n aa t a 

—3226 (Fort Beaufort): Adelaide (-CB), Van Eden NBG 1444131 (BOL): Baddaford 
farm (-DCL Killick 871 (PRE); Near top of Koonop Heights, Schonland 3/09 



(oAM), t-CKion & z^eyner lria iuj id, v - - / > , ~, lf) 

(GRA); Enon, Drege 3489a (K); In collibus prope Uitenhage (-CD), Schlechter 2i 


394 


Journal of South African Botany 


(BOL, GRA, G): Elandsrivier catchment basis, Scharf 1311 (PRE); Addo (-DA/-B), 
Brynard 353 (PRE), Urton 246 (GRA); In der Kloof, Adow, Drege 3489a (SAM); 
Port Elizabeth (-DC), Meyer PRE 36611; 15 mis E of Port Elizabeth, Lewis 4433 
(PRE, SAM); Redhouse, Paterson 1011 (GRA); Mr. Crew’s farm, Coega, De Vos 
2442 (STE), Dahlstrand 665 (PRE). 

— 3326 (Grahamstown): Grahamstown district, Bayliss 1227 (B), Rogers 27741 
(STE); Helspoort (-AB), Bayliss 5722 (C, MO); Alicedale (-AC), Cruden 220 
(STE); Ca. 15 mis SW of Grahamstown (-AD), Lewis SAM 65866 (PRE); 11 mis 
from Grahamstown on Salem road, Cheadle 734 (PRE); Grahamstown (-BC), Daly 
& Sole 67 (PRE), Rogers 27741 (GRA); Grahamstown flats, Britten A60 (GRA); 
Peddie road 15 mis from Grahamstown (-BD), Dyer 1382 (GRA, PRE); De Bega 
valley, Alexandria (-CB), Galpin 10662 (PRE), Galpin s.n. BOL 31761, Archibald 
4118 (PRE); Alexandria, MacOwan s.n., 3/1884 (K), Theron 259 (PRE); Boesmans- 
rivier valley between Grahamstown and Alexandria, De Vos 2443 (STE); Bathurst 
(-DB), Sidey 3673 (PRE). 

— 3327 (Peddie): Keiskamma, Hutton s.n., 3/68 (K). 

Without precise locality: Thom. s.n. (K); Harvey s.n. (K). 

TANZANIA — Iringa, Sao Hill, Watermeyer 42 (K); 11 mis SE of Iringa, Polhill 
& Paulo 1388 (PRE, B, K, SRGH); Songea distr.. Hay 110 (B, K); Njombe distr., 
Leedal 5308 (K); Ufipa distr., Bullock 2248 (K); Kondoa distr., Burtt 961 (K); Ger- 
man East Africa, Speke & Grant s.n. (K). 

ZAMBIA — Abercorn distr., Kawimbe, Richards 8002 (K); Kapatamoyo near 
Fort Jameson, Robson 1040 (K, SRGH); Near Lundazi, Anton-Smith 140 (SRGH). 

MALAWI — Mzuzu distr., Vipya Range, Richards 14458 (K); Mzimba distr., Vi- 
pya plateau, Pawek 8903 (K); 37 mis SW of Mzuzu, Pawek 8050 (SRGH); Nkhata 
Bay distr., Vipya mts., Williamson & Ball 1810 (SRGH); Luwawa, Vipya, Chapman 
1547 (SRGH); Dzalanyama Forest, Adlard SRGH 135188 ; Dedza distr., Ciwan Hill, 
Banda 411 (SRGH); Zomba mts., Whellan 1161 (SRGH); Dedza district, Chongoni 
Forest Reserve, Brummit 10152 (K); Mpata mlonde Hill, Dedza distr., Jackson 2161 
(K); Sangono Hill, Adlard 238A (K). 


Herbarium material of this species from Central East Africa is indistin- 
guishable from the eastern Cape Province material. Baker (1875) also con- 
sidered this to be the case, as he described the only Central African collec- 
tion then known, which had been brought back by Speke and Grant’s 
expedition of 1860, as Montbretia laxifolia Klatt. He did not indicate the 
presence of calli either in his quite extensive description or in the accompa- 
nying figure. However, the specimen cited, preserved in the Kew herbari- 
um, distinctly shows high calli on the perianth. As Baker had omitted to in- 
dicate calli, Klatt thought these Central African plants to be a different 
species and in 1895 he named it T. bakeri, which is a later homonym. 

T. laxifolia and the closely related T. securigera are a vicarious species 
pair occupying neighbouring areas in the Cape Province. They resemble 
each other extensively and are best distinguished in herbaria by their differ- 
ent flowering periods, T. laxifolia being the only autumn flowering species of 
the southern Cape Province. The Cape T. laxifolia also differs from T. secu- 
rigera in its slightly smaller flowers showing a cup whitish inside with slender 
dark lines, higher and narrower, almost peg-like, calli, and anthers reaching 


395 


African genus Tritonia Ker-Gawler: part 2 

slightly higher up the perianth; also in its longer, ellipsoid, erect capsules 
and its usually longer, somewhat more flaccid leaves, as well as in its more 
eastern distribution in the Cape Province. 

It differs from T. nelsonii of the Transvaal in its usually shorter and 
wider leaves with less prominent, more widely spaced veins, in its perianth 
tube which is narrowly tubular in the lower half and in its longer, ellipsoid 
capsules. Three collections from the Central Region of Malawi in the Kew 
herbarium, namely Brummit 10152 and Jackson 2161, both from the Dedza 
district, and Adlard 238A from Sangono Hill, have narrower leaves resem- 
bling T. nelsonii in this feature. The first-mentioned was found on rocky 
slopes and the second in very shallow soil overlying rocks. These habitats 
may perhaps have contributed to their narrower leaves. 

20. Tritonia watermeyeri L. Bol. in Ann. Bol. Herb. 4: 44 (1926). Type: 
Cape, allegedly Vanrhynsdorp, Watermeyer BOL 18050 (BOL, holo-). 

Montbretia watermeyeri (L.Bol.) Sealy in Curtis’s bot. Mag. 174 t. 428 
(1963); Pike in Gard. Chron. 152: 335 (1962). 

leones: Curtis’s bot. Mag. 174 t. 428; a coloured figure by M. Page in 
BOL; this work Fig. 25. 

Plants (100-)200-300 mm long. Corm ovoid, 7-15 mm diam.; tunic 
fibres fine to strong, reticulate. Scape (60-)120-300 mm long, erect or some- 
what flexuose, with old leaf bases up to 80 mm long around its base. Leaves 
3-7, linear-lanceolate, acute, 50— 125(— 150) x 2-8 mm, suberect or spread- 
ing, with strongly undulate margins, sometimes spirally twisted, with a slen- 
der middle vein. Spike simple, distichous, lax, with 2-6 flowers, rarely re- 
duced to one. Bracts membranous, ovate, concave, 6-9 mm long, brown 
and minutely stippled in the upper part, often purplish in the lower; outer 3- 
toothed or irregularly toothed, with purplish veins; inner bidentate, 2- 
veined, with wide, pale, membranous margins. Flowers almost 2-lipped, 
(25-)30-40 mm long, faintly fragrant, burnt orange or pinkish-orange (RHS 
31 A), the upper perianth segments often pale salmon or buff-coloured in- 
side, the lower segments with a red-margined yellow zone in the throat. 
Perianth tube tubular in the lower 3-7 mm, widened to ca. 7 mm diam. at 
the top, 10-16 mm long; segments unequal, obtuse, median posticous seg- 
ment obovate, erect, concave, 15-22 x (8— )12 — 15 mm, lateral posticous seg- 
ments spreading, oblong, 15-18 x 7-10 mm, anticous segments oblanceo- 
late, deflexed just above the base, 11-14 x 5-9 mm, each with a yellow 
truncate callus 3-4 mm high. Stamens curved against the posticous perianth 
segment; filaments 12-18 mm; anthers 4-5 mm long, straight or curved, con- 
tiguous, violet or mauve; pollen pale. Style 17-25 mm; stigmatic branches 
3-4 mm, as high as the anthers or just overtopping them. Capsules obovoid- 


396 


Journal of South African Botany 



Fig. 25 

Tritonia watermeyeri (Malan 112): a, plant and front view of flower; b, outer (left) 
and inner (right) bract; c, flower half in a median cut; d, capsule. 


trigonous, ca. 10 mm long; seeds round, brown, minutely reticulate-collicu- 
late, 1,5 mm diam. 

Flowering period: August to October. 

Distribution: in a small area hardly 60 km in diameter in the Little Karoo 
east of Montagu, to Barrydale (Fig. 21). The Vanrhynsdorp locality is prob- 
ably incorrect. 

CAPE — 3320 (Montagu): Allemorgens (-CB), Malan 112 (NBG, STE); 5,5 mis from 
Montagu towards Ladismith via Kareevlakte (-CC), Barnard BOL 24788 , SAM 
68461 ; 10 mis from Montagu towards Ladismith (-CD), Hurley & Neil BOL 20177, 


African genus Tritonia Ker-Gawler: part 2 397 

Lewis BOL 21648 (BOL, PRE); Kleinvlakte near Barrydale (-DC), Van Breda 4379 
(PRE, K). 

Without precise locality: Ex hort. Barnard , NBG 10167] “Probably sent by 
Watermeyer from Vanrhynsdorp”, BOL 18050. 3 

T. watermeyeri consists of small plants with flowers more or less resem- 
bling those of T. securigera, but differing in the perianth having a somewhat 
more two-lipped appearance. The distinctive character of T. watermeyeri is 
its rather flaccid, often spirally twisted leaves which have undulate margins. 

The species is not closely related to T. tugweiliae or to T. undulata both 
of which have leaves with undulate margins, but have flowers without calli. 
These therefore belong to other sections of the genus. 

21. Tritonia drakensbergensis De Vos, sp. nov. 

Caulus 250-300(-500) mm longus. Folia 4-5, linearia, acuminata, erec- 
ta, 150-250(-300) x 1 ,5— 3( — 4) mm, nervo valido in medio et prope mar- 
gines praedita. Spica subsecunda, floribus (2-)3-7. Bracteae fere chartaceae, 
7-11 mm longae, cinnamomeae; bractea exterior subobtusa vel irregulariter 
dentata vel 3-dentata, nervo medio prominenti; interior basin versus pallida, 
binervata, bidentata, dentibus acuminatis, marginibus latis chartaceis. Flores 
zygomorphi, (22-)25-32 mm longi, miniati. Tubus perianthii late infundibu- 
laris, 8-10 mm longus, brevior segmentis; segmenta longitudine subaequalia, 
15-20 x ca. 5-8 mm, segmento postico parum latiore, segmentis anticis tri- 
bus in fauce callo flavo subtriangulari praeditis. Stamina segmentum posti- 
cum versus posita; filamenta ca. 9-12 mm longa; antherae 5-6 mm longae, 
contiguae, infra vel supra medium segmentorum attingentes. Stylus 14-18 
mm longus, ramis stigmaticis ad vel supra apices antherarum attingentibus. 

Type: Cape, Transkei, Little Bush, Cala, 4 000', Pegler 1699 (BOL, holo-). 

Plants 250-300(-500) mm long. Corm not seen. Scape 250-300(-450) 
mm long, simple, erect or slightly curved. Leaves 4-5, linear, acuminate, 
rather rigid, erect, 150-250(-300) x 1 ,5— 3( — 4) mm, with a strong middle 
vein and a prominent vein near each edge (dried state), almost reaching the 
base of the spike. Spike subsecund, lax, with (2-)3-7 flowers. Bracts almost 
papery, golden-brown, 7-11 mm long; outer subobtuse or irregularly 
toothed or 3-toothed, with a prominent median vein; inner pale towards the 
base, bidentate, 2-veined, with acuminate teeth and wide papery margins. 
Flowers zygomorphic, (22-)25-32 mm long, red or pinky-scarlet. Perianth 
tube widely funnel-shaped, 8-10 mm long, shorter than the segments, seg- 
ments subequal in length, 15-20 x ca. 5-8 mm, with the posticous segment 
widest, and the three anticous segments with an almost triangular yellow cal- 


398 


Journal of South African Botany 


lus in the throat. Stamens and style curved towards the posticous segment; 
filaments ca. 9-12 mm long; anthers contiguous, 5-6 mm long, curved, 
reaching less than, to more than halfway up the perianth segments. Style 
14-18 mm; stigmatic branches 2,5 mm long, at or above the anther tips. 
Capsules not seen. 

Flowering period: January to early March. 

Distribution: on the southern parts of the Drakensberg range in the 
north eastern Cape Province and Transkei, at altitudes of 1 100 to 2 300 m 
(Fig. 22). 

TRANSKEI — 3127 (Lady Frere): In monte Kwenkwe in ditione Maclear (-DA), 
H. Bolus s.n. , Jan. 1896 (BOL); Little Bush, Cala, Pegler 1699 (BOL). 

— 3128 (Umtata): Drakensberg, on Tsitsu footpath, distr. Maclear, Galpin 6849 
(BOL, GRA, PRE, K); Engcobo Mountain (-CA), Flanagan 2659 (BOL). 

CAPE — 3028 (Matatiele): Naudesnek (-CA), Marais 1374 (PRE). 

— 3127 (Lady Frere): Mountain 2,5 mis NE of Barkley Pass (-BB), Nordenstam 2009 
(S, PRE). 

This species is distinguished by red (Galpin) or pinkish-scarlet 
(Nordenstam) flowers with a wide funnel-shaped perianth tube about half 
the length of the segments and with a yellow callus on each of the three anti- 
cous segments; and by its long, linear, rather rigid erect leaves with three 
prominent veins, two of which are near the leaf edges. 

It is closely related to the laxifolia - nelsonii - strictifolia complex, resem- 
bling the narrow-leaved specimens of T. nelsonii in habit. It differs from all 
three species of the complex in its shorter perianth tube; from T. laxifolia 
and T. nelsonii also in its leaves with prominent near-marginal veins and 
flowers with shorter calli; and from T. strictifolia in its much longer, linear 
leaves and different flowering period. 

22. Tritonia nelsonii Bak. Handb. Irid. 195 (1892) & in FI. Cap. 6: 127 
(1896); Verdoorn, in Flow. PI. Afr. 33 pi. 1315 (1959). Type: Transvaal, 
Houtbosberg mountain summit. Nelson 440 (K, holo-; PRE, iso-). 

T. petrophila Bak. in Bull Herb. Boiss. Ser. 2, 4: 1004 (1904). Type: 
Transvaal, Granitfelsen bei Modderfontein, Conrath 585 (Z, holo-; K). 

Icon.: Flow. PI. Afr. 33 pi. 1315; Letty pi. 38 (1962). 

Plants (250-)500-900 mm long. Corm 15-30 mm diam., ovoid; tunic 
fibres fine, reticulate in the upper half, parallel in the lower. Scape (250-) 
500-900 mm long, slender, simple or with 1-2 branches. Leaves 4-8, sub- 
erect or slightly spreading, linear, acuminate, firm, striate, (200-)450-900 x 
l,5-4(-6) mm, reaching from below to sometimes above the spike; cauline 
leaves 1-2, slightly shorter. Spike somewhat lax, almost secund, often 1-2- 
branched, with 5-12 or more suberect flowers. Bracts membranous but 


African genus Tritonia Ker-Gawler: part 2 


399 


greenish at the base at first, brown or minutely brown-striped in the upper 
half and along the edges, (8— )10— 15 mm long; outer 3-toothed or acute, 
striate, with a stronger median vein; inner bidentate, 2-veined, shorter than 
the outer. Flowers (22-)25-35(-38) mm long, light flame to bright orange- 
red or bright red, with a high greenish-yellow callus on each anticous peri- 
anth segment. Perianth tube funnel-shaped, (10— )14— 18 mm long, equal to 
or slightly longer or rarely shorter than the segments; segments unequal, ob- 
lanceolate, obtuse, (10— )12— 17 x 5-10 mm, the posticous segment slightly 
hooded, to 10 mm wide, others narrower, spreading or recurved, the three 
anticous segments with a high greenish-yellow callus on a red-margined 
blotch, and with a pale dorsal median band. Stamens curved against the 
posticous segment; filaments 12-14 mm long, white; anthers 5-6 mm long, 
contiguous, with purple lateral lines, reaching the middle or higher up the 
perianth segments; pollen pale yellow. Style 17-25 mm, white; stigmatic 
branches 2-3 mm long, rarely overtopping the anthers. Capsules subglobose 
to shortly ellipsoid, membranous, 7-10 mm long; seeds shiny black, finely 
foveolate-reticulate, 2 mm diam. 

Flowering period: (December to) February to March (to June). 

Distribution: northern Transvaal and throughout the Highveld of Trans- 
vaal (Fig. 22). 


TRANSVAAL— 2229 (Waterpoort): Hangklip-Louis Trichardt (-DD). Breme- 
kamp & Schweickerdt TM 29855 (PRE); Soutpansberg, Dr. Bird’s farm, Meeuse 
9785 (PRE, SRGH, S). 

— 2230 (Messina): Entabeni, Laastgevonden (-CC), Koker 22 (PRE). 

— 2329 (Petersburg): Happy Rest (-BA), Gerstner 6103 (PRE); Louis Trichardt 
(-BB), Breyer 19491 (PRE); Houtbosberg (-DD), Schlechter 4408 (GRA, PRE); 
Near Elaenertsburg (-DD), Prosser 1918 (PRE), Codd 9439 (PRE, K). 

—2330 (Tzaneen): Broederstroom, Letaba (-CD), Leighton 3215 (PRE). 

—2427 (Thabazimbi): Farm Groothoek, western extremity of Waterberg (-BC), 
Codd 3988 (PRE); Witpoortjie Falls (-CA), Davey 59 (NU). 

—2526 (Zeerust): Zeerust (-CA), Jenkins TM 11665 (PRE); Kraalkop (-CD), De 
Wit 957 (PRE); Swartruggens (-DA), Fouche PRE 36618. 

—2527 (Rustenburg): Rustenburg (-CA), Rogers 23436 (K); F n a ™ ^‘ nh ' 11 . n ^ ar 
Rustenburg, Codd 1080 (PRE); Hennops River (-DA), Repton 496 (PRE); Pelinda ' 
ba (-DD), Strey 3046 (PRE); Uitkomst, Rustenburg, Coetzee 724 (PRE) 

—2528 (Pretoria); Kameeldrift road 35 mis NE of Pretoria (-BC), J* e Pj° n - 713 
(PRE); Lynwood (-CA), Marsh PRE 28583 (K), Verdoorn 2416 (PRE); Onderste- 
poort (-CA), Pole Evans 429 (PRE); Near Onderstepoort, Meeuse 9242 (SRGH) 
Near Silverton (-CB), Repton 382 (PRE); Fountains (-CC), Munro PRE 57-9 Ver- 
doorn 723 (PRE); Premier Mine (-DA), Rogers 25369 (PRE); Renosterkop, Bronk- 

horstspruit district (-DB), Young 2149 (PRE, K). , DT?PV luiHHplhimr 

—2529 (Witbank): Loskopdam, Nooigedagt (-AD), Theron 1212 (PRE), Midde g 

(-CD), King s.n. (C). , 

— 2530 (Lydenburg): Waterval Boven (-CB), Taylor 196 7 (PR b NW * 

—2626 (Klerksdorp): Lichtenburg (-AA), Jenkins TM 11478 (PRE), 1- m 


400 


Journal of South African Botany 


Ventersdorp (-BA), Codd 2660 (PRE), Sidey 97 (S); Goedgedacht, Ventersdorp 
(-BD), Sutton 567 (PRE). 

— 2627 (Potchefstroom): 6 mis N of Fochville (-AD), Dahlstrand 869 (C); Near Kru- 
gerdorp (-BB), Meeuse 9060 (L); Little Falls, Roodepoort Ridge (-BB), Mogg 20233 
(PRE). 

— 2628 (Johannesburg): Modderfontein (-AA), Conrath 585 (K, Z); Near Johannes- 
burg, Robinson 57 (NU), Tucker s.n. (BOL), Rogers 19234 (NH); Weltevreden 10 
mis NW of Johannesburg, Mogg 22674A (PRE); Zoo Kopjes, Moss 8604 (BM, K, 
PRE); Heidelberg (-AD), Leenderts 1024 (BOL), Burtt-Davy 3098 (PRE), Strey 
3065 (PRE); Heidelberg Kloof (-AD), Mogg 20500 (PRE); Suikerbosrand Nature 
Reserve (-AD/-CA), Kroon 189 (PRE), Bredenkamp 441 (PRE). 

Without precise locality: Bechuanaland, Pole-Evans & Ehrens 1915 (K, SRGH). 

This species is closely related to T. laxifolia, differing from this in its fun- 
nel-shaped perianth tube which widens from close to its base, and in its 
longer, narrower, linear leaves narrowing gradually to an acuminate tip. The 
veins are closely spaced and the leaves appear striate in the dried condition. 
Leaves in shade are less striate. The calli on the anticous perianth segments 
resemble those of T. laxifolia. 

Two growth forms occur which were formerly treated as distinct species. 
In moister regions the plants are somewhat lax with fewer and larger flow- 
ers; in dryer areas they are more erect and have smaller flowers (Verdoorn, 
1959). 

23. Tritonia atrorubens (N.E.Br.) L. Bol in J. Bot. Lond. 67: 133 (1929). 

Gladiolus atrorubens N.E.Br. in Kew Bull. 1914: 135 (1914). Type: 
Transkei, Kentani, near Manubi, Saxton 13 (K, holo-). 

T. flanaganii L. Bol. in Ann. Bol. Herb. 2: 161 (1918). Syntypes: Cape, 
near Impetu between Komga & Kei Mouth, Flanagan 55 (BOL, lecto-); 
Transkei, flats near Kentani, Pegler 392 (BOL, GRA). 

Plants 300-600 mm long. Corm globose, 10 mm diam.; tunic fibres re- 
ticulate in upper half, parallel towards the base. Scape 300-600 mm long, 
erect, slender, simple. Leaves 3-6, linear, erect, acuminate, striate with 
three or more strong veins, 150-400 x (l-)2-5 mm, reaching or almost 
reaching the base of the spike; cauline leaves 2-4, 100-200 mm long, subu- 
late. Spike somewhat secund, rather lax, with 2-9 flowers. Bracts membra- 
nous, 8-10 mm long, brown in the upper half; outer acute or acuminate or 
irregularly toothed, veined, the median vein brown upwards; inner slightly 
shorter, bidentate, 2-veined, with wide, pale, subhyaline margins. Flowers 
obliquely funnel-shaped, 25-30 mm long, dark red, drying to a blackish 
purple-red. Perianth tube 15-17 mm long, funnel-shaped, almost straight, 
1,5 to almost 2 times the length of the segments; segments subequal, oblong, 
obtuse, 8-12 x 4-8 mm, the posticous segment wider than the others, the 


401 


African genus Tritonia Ker-Gawler: part 2 

anticous segments each with a high callus, with that on the median segment 
larger than on the lateral ones and yellow. Filaments 10-11 mm, reddish; 
anthers 5—7 mm long, dark, reaching more than halfway up the perianth seg- 
ments. Style 18 mm, reddish; stigmatic branches 3-4 mm long, reaching the 
anther tips. 

Flowering period; December to January. 

Distribution: on a small area hardly 50 km in extent, west and east of the 
Cape-Transkeian border, at low altitude (Fig. 22). 

CAPE-TRANSKEI — 3228 (Butterworth): Kentani, near Manubi (-BC), Saxton 
13 (K); Near Mpetu (-CA), Flanagan 55 (Bol, PRE); Near Kentani (-CB), Peeler 
392 (BOL, GRA, PRE). 

Without precise locality: Marloth 8624 (PRE); Transkeian Country, Bowker 207 
(K). 

This rare species, found only a few times near the southern part of the 
Cape-Transkeian border, is distinguished by its long, narrow, linear leaves 
with a structure almost like that of T. Bakeri , by its outer bracts often with acute 
or acuminate tip, and its dark red flowers which, on drying out, become red- 
dish-black, and which have a perianth tube longer than the segments and 
calli on the lower segments. The callus on the median anticous segment is 
yellow and larger than the lateral ones which seem to be dark in colour 
(red?). 

24. Tritonia moggii Oberm. in Kirkia 3: 24 (1963). Type: Mozambique, 
Inhaca Island, Ridge Woodland SW of Saco Bay, Mogg 27315 (PRE, 
holo-; J, SRGH). 

leones: Umpublished, by M. Walgate, in BOL; this work Fig. 26. 

Plants (200-)300-500 mm long. Corm 10-15 mm diam., globose, some- 
times with small cormlets at the ends of 20-30 mm long stolons; tunic fibres 
fine, elongated-reticulate. Scape erect, simple or with 1-3 branches, 
(200-)300-500 mm long. Leaves 6-9, linear-lanceolate, acuminate, soft tex- 
tured, 150-400(-500) x 4— 8(— 11) mm, sometimes overtopping the spike, 
with a prominent middle vein and a prominent lateral vein almost on each 
leaf edge; cauline leaves shorter and narrower. Spike almost secund, lax, 
simple or few-branched, with 8-10 or more flowers. Bracts membranous, 
papery, 6— 12(— 15) mm long, brown or reddish-brown in the upper half, with 
acuminate teeth; outer finely veined, acute or 2-3-toothed, with the median 
tooth often reduced; inner bidentate, 2-veined, equal to or sligh shorter 
than the outer. Flowers (40-)45-55 mm long, orange-pink or salmon-pink, 
with a yellow, red-margined blotch on the three anticous perianth segments. 


402 


Journal of South African Botany 



African genus Tritonia Ker-Gawler: part 2 


403 


Perianth tube (24-)30-35(-40) mm long, narrowly tubular for most of its 
length, widened slightly near the top; segments oblong-lanceolate, obtuse 
(12— )15— 22 x 6-9 mm, the median posticous segment slightly wider, the 
median anticous, and sometimes also one or both lateral anticous segments 
with a yellow callus. Filaments 12-16 mm; anthers 6-7 mm long, curved, 
reaching halfway or higher up the perianth segments. Style 33-40 mm; stig- 
matic branches ca. 4 mm long, reaching or overtopping the anthers. Cap- 
sules ellipsoid or turbinate, membranous, 8-15 mm long; seeds round, shiny 
black, finely reticulate-foveate, ca. 2 mm diam. 

Flowering period: April to July, rarely to October. 

Distribution: Mozambique, coastal regions and Inhaca Island, as well as 
in the inland district of Gaza (Fig. 22). 

MOZAMBIQUE — Gaza, between Manjacaze & the Coolelo monument, Lemos 
& Balsinhas 115 (PRE); Masieni, Van Dam PRE 58376 , TM 25358 (PRE); Chibuto, 
8 km depois de Maniquenique, Lemos & Balsinhas 138 (PRE); Maniquenique Expt. 
Esta^ao, Barbosa & De Lemos 7599 (SRGH); Chicumbane, Junod 355 (PRE); Sul 
do Save, Pedro & Pedrogao 1331 (PRE, K); Quelimane, Sim 20507 (PRE); Inham- 
bane, Gomes & Sousa 1876 (K); Inhaca Island, Barbosa 7630 (PRE, K), Mogg 27315 
(PRE, SRGH), Weintraub J20439 (BOL, SRGH), Hancock J28015 (PRE); 100 mis 
NW of Lorenzo Marques (Maputo), Debenah s'.n. (STE). 


This species is readily distinguished by its somewhat zygomorphic flowers 
with a perianth tube considerably longer than the segments and usually a 
single callus, and by its leaves that have a prominent vein almost on each 
leaf margin. It stands nearest T. laxifolia and T. nelsonii which have a shorter 
perianth tube and three calli. Several collections seem to connect the species 
with T. laxifolia, e.g., Van Dam s.n. (PRE 58376), Lemos & Balsinhas 138 
(PRE) and Barbosa & De Lemos 7599 (SRGH). These show a small callus 
also on one or both lateral anticous perianth segments. These plants have 
perhaps incorporated some genes of T. laxifolia. They are from the Gaza 
district of Mozambique which is near the border with Malawi where 
T. laxifolia occurs. 

Some of the flowers of the Pedro & Pedrogao collection no. 1331 are 
only 35 mm long with a perianth tube of 18 mm. 

25. Tritonia delpierrei De Vos, sp. nov. 

F'g- 27. 

Plantae 230-350 mm longae. Cormus ovoideus, 12-25 mm diam., fibris 
tunicae validis, elongato-reticulatis, basin versus subparallelis. Caulis rigidu- 
lus, aliquantum flexuosus, suberectus, 230-350 mm longus, base collo folio- 


Fig. 26 

Tritonia moggii (from a drawing by M. Walgate in BOL): a, plant, b, outer 
inner (right) bract; c, flower, front view; d, capsules; p, posticous penantn 


(left) and 
segment. 


404 


Journal of South African Botany 


rum veterum circumnexa, 1— 3-ramosus. Folia 3—5, linearia, acuminata, sub- 
succulenta, 120-300 x 2-3(^l) mm, suberecta vel leviter patula, subglauca. 
Spica laxa, subsecunda vel interdum subdisticha, ramo principali 5-9-floren- 
ti. Bracteae in dimidio superiore membranaceae, minute sphacelatae, apici- 
bus brunneis, in dimidio inferiore viridulae, 5-10 mm longae; bractea exte- 
rior minute 2-3-dentata vel subacuta vel subobtusa; interior interdum parum 
brevior quam exterior, bidentata, binervis. Flores subzygomorphi, in medio 
diei fere hypocrateriformes, 22-25(-28) mm longi, eburnei et flavidi, leviter 
fragrantes. Tubus perianthii tubularis, 12-18 mm longus, 1,5-2-plo longior 
quam segmenta, in dimidio superiore gradatum dilatatus ad 2 mm diam.; 
segmenta 8— 10(— 12) mm longa, patentia, obtusa, segmentis posticus ellipti- 
cis, eburneis, 3-5 mm latis, segmentis anticis oblanceolatis, flavidis, 3-4 mm 
latis, in quoque callo flavido ad 1 mm alto. Stamina et stylus parum segmen- 
tum posticum versus positi, eburnei; filamenta 5-7 mm longa; antherae con- 
tiguae, 3-5 mm longae, supra dimidium segmentorum attingentes. Stylus 
18-25 mm longus, ramis stigmaticus 1,5-2 mm longis, supra apices anthera- 
rum attingentibus. Capsulae subrotundo-trilobatae, 5-10 mm longae, semi- 
nibus protuberantibus, globosis, brunneis, minute reticulatis, 2 mm diam. 

Type: Cape, southern Richtersveld, top Helskloof, De Vos 2462 (STE, holo-). 

Plants 230-350 mm long. Corm ovoid, 12-25 mm diam.; tunic fibres 
strong, elongate-reticulate, almost parallel towards the base. Scape rather 
rigid, slightly flexuose, suberect, 230-350 mm long, with a collar of old leaf 
bases up to 60 mm long around the base, 1-3-branched, the branches some- 
what divergent at their bases, then ascending. Leaves 3-5, linear, acumi- 
nate, slightly succulent, 120-300 x 2-3(-4) mm, suberect or slightly spread- 
ing, lightly glaucous, with a strong middle vein or sometimes two strong 
veins; cauline leaves 1-3, 10-60 mm long. Spike lax, subsecund or some- 
times subdistichous, with 5-9 flowers on the main branch. Bracts membra- 
nous and minutely sphacelate in the upper half, with brown tips, greenish in 
the lower half, 5-10 mm long; outer minutely 2-3-dentate, or shortly suba- 
cute or subpbtuse; inner sometimes slightly shorter than the outer, 2-veined, 
2-dentate with acuminate teeth. Flowers slightly zygomorphic, sometimes in- 
verted, almost salver-shaped in the middle of the day with spreading seg- 
ments, 22-25(-28) mm long, cream and pale yellow, sometimes faintly 
flushed with purplish-pink on drying, faintly fragrant. Perianth tube tubular, 
12-18 mm long, 1,5-2 times longer than the segments, gradually widened in 
the upper half to 2 mm diam.; segments (8— )10— 12 mm long, spreading, ob- 
tuse, posticous segments elliptical, cream, 3-5 mm wide; anticous segments 
oblanceolate, pale yellow, 3-4 mm wide, each with a small yellow callus up 
to 1 mm high .on the lower half. Stamens and style placed nearer the posti- 


African genus Tritonia Ker-Gawler: part 2 


405 



406 


Journal of South African Botany 


cous side of the flower, cream; filaments 5-7 mm long, inserted near the top 
of the perianth tube; anthers contiguous, 3-5 mm long, reaching above the 
middle of the perianth segments, with pale yellow pollen. Style 18-25 mm; 
stigmatic branches 1,5-2 mm long, slightly expanded at their tips, overtop- 
ping the anthers. Capsules subrotund-trilobed, 5-10 mm long, with bulging 
seeds; seeds globose, brown, finely reticulate, 2 mm diam. Chromosome 
number 2n = 20 (f. Goldblatt). 

Flowering period: late July to early October, probably depending on the 
beginning of the rainy season. 

Distribution: southern Richtersveld, Namaqualand (Fig. 22). 

CAPE — 2817 (Vioolsdrif): Helskloof, between Stinkfontein and Orange River 
(-CD), Delpierre s.n., Sept. 1977 (MO); Top Helskloof, between Modderdrif and 
Eksteensfontein, Delpierre s.n. NBG 118049\ Cultivated from seed, Goldblatt 4984 
(MO); Top Helskloof, De Vos 2453 (STE); Cultivated from corms, De Vos 2462 
(STE). 

The range of this new species, as well as that of T. marlothii , is far dis- 
tant from other species of Tritonia, their nearest neighbours being 
T. karooica and T. kamisbergensis. Possibly T. delpierrei stands nearest 
T. karooica which also has yellow, only slightly zygomorphic flowers with a 
perianth tube longer than the segments and calli on the lower segments, as 
well as capsules with large, bulging seeds. They differ, however, in their 
general habit, T. delpierrei being a more slender plant with narrower, more 
erect, slightly succulent leaves and paler flowers which become almost sal- 
ver-shaped in the middle of a hot day and which have smaller, spreading, 
but not reflexed, segments. In late afternoon the flowers close somewhat, 
the three lower segments becoming grouped closer together. This gives the 
flowers a somewhat two-lipped appearance, more or less resembling other 
species of section Montbretia. 

The species was discovered by Prof. G. R. Delpierre, whence the name. 
It seems to be quite common in the southern part of the Richtersveld to- 
wards and at the top of Helskloof. 


26. Tritonia marlothii De Vos, sp. nov. 

Fig. 27. 

T. delpierrei affinis, foliis et habitu similaribus; differt imprimis floribus 
et stylo longioribus et tubo perianthii 3-4-plo longiore quam segmentis. 

Type: Cape, Richtersveld, Marloth 12296 (BOL, holo-; PRE). 

Plants 200-350 mm long. Corm not seen. Scape rather rigid, slightly flex- 
uose, suberect, 200-350 mm long, 2-3-branched. Leaves 4-5, linear, acumi- 


407 


African genus Tritonia Ker-Gawler: part 2 

nate, slightly succulent, 100-240 x l,5-3(-4) mm, with one or two promi- 
nent veins; cauline leaves 1-2, 30-50 mm long. Spike distichous, rather lax, 
with 6-12 flowers on the main branch. Bracts membranous, almost papery 
and finely sphacelate in the upper half, with dark brown tips, 7-10 mm long; 
outer obtuse or minutely 2-3-toothed; inner bidentate with short acuminate 
teeth. Flowers slightly zygomorphic, somewhat salver-shaped, 35-45 mm 
long, apparently pink. Perianth tube tubular, slightly curved, 28-35 mm 
long, 3-4 times the length of the segments, gradually widened in the upper 
half to 2 mm diam.; segments oblanceolate or oblong, obtuse, spreading, 
8-10 x 2,5—4 mm, with a small yellow callus ca. 1 mm high on the lower 
half of each anticous segment. Filaments 6-8 mm long; inserted near the top 
of the perianth tube; anthers contiguous, 4 mm long, reaching above the 
middle of the segments. Style 32-40 mm; stigmatic branches 1,5-2 mm long, 
overtopping the anthers. Capsules subrotund-trigonous or shortly obovoid- 
trigonous, 6-10 mm long; seeds subglobose, dark brown, 2 mm diam., min- 
utely reticulate-foveolate. 

Flowering period: late August to September. 

Distribution: southern Richtersveld, Namaqualand (Fig. 22). 

CAPE— 2817 (Vioolsdrif): Brakfontein (-CC), Marloth 12296 (BOL, PRE). 

This new species, found only once in 1925, is very closely related to 
T. delpierrei, being similar to the latter except for its longer flowers, longer 
perianth tube and style, and perhaps different flower colouring. The flower 
colour of the dried specimens have been noted on the herbarium sheet in 
PRE as “apparently pink” and in BOL as “flowers appear purple to pink, 
pale”. The note in BOL is not in Marloth’s handwriting. There is a possibil- 
ity that the flowers may be cream or pale yellow, like those of T. delpierrei 
which sometimes tend to become flushed with purplish-pink on drying out. 
Whether the bracts are greenish in their lower half as in T. delpierrei, could 
not be determined. 

As no intermediates between the two taxa have as yet been found, they 
are here treated as separate species. 

27. Tritonia karooica De Vos, sp. nov. 

Fig. 28. 

Plantae 100-200(-230) mm longae. Cormus ovoideus, basi subcomplana- 
tus, 10-25 mm diam., fibris tunicae validis, superne elongato-reticulatis, ba- 
sin versus subparallelis. Scapus 40-150 mm longus, interdum 1-ramosus, cir- 
ca basin fibris veteribus. Folia 6-10, disticha, flabellata, lanceolato-falcata, 
reflexa, percurvata vel expansa, 40-100(-150) x 3-7 mm. Spica flexuosa, 
floribus 2-6 vel raro pluribus. Bracteae membranaceae, ( 12—) 1 5—2.0 mm 


408 


Journal of South African Botany 


longae; bractea exterior acuta v. 3-dentata, apicem versus subtiliter sphace- 
lata, porphyreo-nervi; interior bidentata, binervis. Flores leviter zygomorphi, 
infundibulares, demum segmentis perianthii reflexis, 40-60 mm longi, ple- 
rumque ochracei, armeniaco suffusi, fusco venosi, fragrantes. Tubus peri- 
anthii 25-32 mm longus, infundibularis; segmenta obovato-spathulata, obtusa 
v. subobtusa, 12-18 mm longa, longitudine subaequalia, segmenta postica 
7-12 mm lata, mediano latissimo, segmenta antica 6-7 mm lata, in quoque 
callo luteo vix 1 mm alto. Stamina segmentum posticum versus curvata; 
filamenta 15-22 mm longa; anther ae curvatae, fere dimidium segmentorum 
attingentes. Stylus 28-35 mm longus; rami stigmatici antheras attingentes v. 
parum superantes. Capsulae breviter ellipsoideo-trigonae v. obovoideo-tri- 
gonae, seminibus leviter protuberantibus, globosis, brunneis, ca. 2 mm 
diam., subtiliter reticulato-foveatis. 



Fig. 28 

Tritonia karooica (De Vos 2259, 2458 ): a, plant; b, outer (left) and inner (right) 

bract; c, capsules. 


African genus Tritonia Ker-Gawler: part 2 


409 


Type: Cape, Murraysburg, Tyson 311 (BOL, holo-; NH, SAM, BM, K). 

Plants 100-200(-230) mm long. Corm ovoid, somewhat flattened at the 
base, 10-25 mm diam.; tunic fibres strong, elongated-reticulate above, at 
the base subparallel and often somewhat plate-like. Scape 40-150 mm long, 
sometimes with one branch, with a short collar of old fibres around its base. 
Leaves 6-10, lanceolate-falcate, acute or obtuse, in a dense, fan-shaped dis- 
tichous basal group, usually strongly curved, reflexed or spreading, 
40-100(-150) x 3-7 mm, with a prominent middle vein. Spike distichous, 
flexuose, with 2-6 or rarely more flowers. Bracts membranous, papery, 
(1 2-) 15-20 mm long; outer acute or 3-toothed with the middle tooth longest, 
finely sphacelate towards the tip, with reddish-brown veins; inner bidentate, 
2-veined. Flowers slightly zygomorphic, funnel-shaped, ultimately with re- 
flexed perianth segments, 40-60 mm long, usually dull yellow or brownish- 
yellow or sometimes cream, orange or salmon, flushed with orange or pink 
outside, conspicuously dark-veined, fragrant in the evening. Perianth tube 
25-32 mm long, funnel-shaped, narrow in the lower half, widened gradually 
in the upper half to ca. 8-10 mm diam; segments obovate-spathulate, obtuse 
or subobtuse, 12-18 mm long, subequal in length, upper segments 7-12 mm 
wide, the median one widest, anticous segments 6-7 mm wide, each with a 
yellow callus 6-10 mm long and hardly more than 1 mm high. Stamens 
curved towards the posticous segment; filaments 15-22 mm; anthers 6-7 mm 
long curved, reaching about halfway up the perianth segments. Style 28-35 
mm; stigmatic branches 4-5 mm long, reaching or slightly overtopping the 
anthers. Capsules shortly ellipsoid-trigonous or obovoid-trigonous, with 
bulging seeds, 8-12 mm long; seeds globose, brown, finely reticulate- 
foveate, ca. 2 mm diam. 

Flowering period: August to September, sometimes to October. 

Distribution: widespread throughout the western and north western parts 
of the Great Karoo, mostly above the Nuweveld escarpment (Fig. 22). 


CAPE— 2918 (Gamoep): Areb (-AC), Barker 8364 (NBG, PRE), Lewis SAM 
65405 (PRE); About 30 mis NE of Springbok (-CB), Lewis 3501 (SAM); 20 mis NE 
of Springbok (-CA), Barker 6670 (BOL, NBG). 

—2919 (Pofadder): 5 mis NE of Pofadder (-AB), Lewis SAM 65406 (PRE); Between 
Bladgrond and Pofadder (-BA), Barker 8340 (NBG, STE). , 00/13 

—2922 (Prieska): Prieska (-DA), Bryant 1140 (BOL, PRE, K), Pole-Evans 18803 
(PRE) 

—2923 (Douglas): Strydenburg (-DC), Schweickerdt 2604 (PRE); South of Stryden- 
burg, Taylor 855 (BOL). . . 

—3018 (Kamiesberg): Dabidas farm, ca. 5 km SE of Aalwynsfontein (-BC), lhomp- 

son 2853 (PRE). , , . , c , 

—3021 (VanWyksvlei): Jasperskop, ca. 20 mis N of VanWyksvlei (-BB), Stay net 

s.n., 21/8/63 (NBG, STE). 

— 3022 (Carnarvon): Carnarvon (-CC), Curry BOL 13682. 


410 


Journal of South African Botany 


— 3118 (VanRhynsdorp): Nuwefontein, fired at Whitehill, Archer 683 (BOL). 

— 3119 (Calvinia): Calvinia (-BD), Hanekom 2114 (PRE), Schmidt 335 (PRE). 

—3120 (Williston): 20 mis E of Calvinia (-AC), Barker 9520 (NBG, STE); 16 mis W 
of Williston (-BC), Acocks 14695 (PRE); Near Williston (-BD), Schlieben 9159 
(PRE); Rietfontein, Hanekom 2114 (MO); Between Middelpos and Calvinia (-CC), 
Goldblatt 566 (BOL). 

— 3121 (Fraserburg): Between Fraserburg and Carnarvon (-BD/-DA), Hafstrom & 
Acocks 280 (PRE, S); Fraserburg-Williston road (-BD), De Vos 2259 (STE). 

— 3123 (Victoria West): Murraysburg (-DD), Tyson 311 (BOL, SAM, BM, K, NH). 
— 3124 (Hanover): Hanover (-AB), Burger s.n., Sept. 1907 (GRA). 

— 3221 (Merweville) : Layton, Leeugamka-Fraserburg road (-BB), Shearing 101 
(PRE); Shearing sub De Vos 2458 (STE). 

Without precise locality: Bushmanland, Bolus 31779. (BOL). 

This very distinct species has been mistakenly identified in South African 
herbaria as T. flava. An examination of the holotype of the latter in BM 
showed that the present species is, except for the yellow colour of the flow- 
ers, totally different. It differs from T. flava in its narrower, usually re- 
curved, distichous leaves spread fan-wise, and in its perianth with dark 
veins, long perianth tube, and long, but low, calli on the lower segments. 

T. karooica is readily distinguished by its usually strongly reflexed leaves 
and large funnel-shaped, usually brownish-yellow flowers often flushed with 
orange or pink outside, with a perianth tube about twice the length of the 
segments, which have five or more dark longitudinal veins, and later become 
strongly reflexed; the calli on the anticous perianth segments are long but 
hardly more than 1 mm in height. 


28. Tritonia florentiae (Marl.) Goldbl. in Bothalia 11: 281 (1974). 

Gladiolus florentiae Marl, in Trans. Roy. Soc. S. Afr. 2: 241 (1912). 
Type: Cape, near Groot Tygerberg, Prince Albert district, Marloth 4452 
(PRE, holo-; BOL). 

Montbretiopsis florentiae (Marl.) L. Bol. in S. Afr. Gard. 19: 215 (1929); 
Phillips, Gen. S. Afr. Flow. PI. ed. 2, 219 (1951). 

leones: Marloth, FI. S. Afr. 4: t. 47a (1915); this work Fig. 29. 

Plants 70-120 mm long. Corm ovoid to pear-shaped, 10-15 mm diam.; 
tunic fibres very fine, elongated-reticulate. Scape very short, hidden by leaf 
sheaths and leaf remains which form a collar ca. 30 mm long around the 
base. Leaves 4-8, basal, distichous, scimitar-shaped, recurved, acute to sub- 
acute, 50-70 x 3-6 mm, with a slender middle vein. Inflorescence with 
2-3(-4) flowers, each terminal on its own peduncle of ca. 10 mm and sub- 
tended by a pair of bracts. Bracts membranous, papery, very thin, soon 
crumpled, 12-18 mm long, pale straw-coloured, finely stippled or striate 
with short brown lines, hidden by the leaf sheaths; outer greenish in the low- 


African genus Tritonia Ker-Gawler: part 2 


411 



Fig. 29 

Tritonia florentiae (De Vos 2282): a, plant and front view of flower; b, flower half in 

a median cut. 


er part, acute, soon lacerated at the tip; inner bidentate. Flowers zygomor- 
phic, (32-)44-50 mm long, bright yellow (RHS 12A, B), reddish in the 
throat and with an arrow-shaped red mark on the lower segments. Perianth 
tube (15-)18-24 mm long, almost straight, tubular in the lower 8-10 mm, 
gradually widened and funnel-shaped upwards; segments unequal, 
(15-)20-25 mm long, obtuse to subacute, median posticous segment largest, 
obovate, 13— 15(— 18) mm wide, slightly concave, sometimes emarginate; lat- 
eral posticous 8-10 mm wide, oblanceolate, anticous segments 5— 7(— 9) mm 
wide, oblanceolate, each with a large, triangular, yellow callus with a crisped 
upper margin. Stamens curved towards the posticous segment filaments 
10-13 mm; anthers 5-6 mm long, contiguous, facing downwards, reaching 
above the middle of the posticous segment, yellow or purple with mauve or 


412 


Journal of South African Botany 


whitish pollen. Style 25-30(-33) mm; stigmatic branches 3 mm long, often 
overtopping the anthers. Capsules inflated, ovoid to shortly ellipsoid, 15-20 
mm long, each on its own stalk; seeds dark brown, round or angled, smooth, 
ca. 2 mm diam. 

Flowering period: May to July. 

Distribution: western and southern parts of the Great Karoo from Cal- 
vinia to Beaufort West and Prince Albert (Fig. 21). 

CAPE — 3119 (Calvinia): Stompiesfontein, Compton 11151 (NBG); South of 
Bloukranspas on road to Ceres (-DC), De Vos 2343 (STE). 

— 3219 (Wuppertal): Ceres district, 10 mis NE of Tulpfontein (-DB), Acocks 14456 
(PRE): Ca. 160 kms S of turnoff to Ceres from Calvinia (-DD), Goldblatt 3921 
(MO). 

— 3222 (Beaufort West): 14,5 mis S Beaufort West (-DA), Acocks 20441 (PRE). 

— 3322 (Oudtshoorn): Prince Albert, N foot of Tierberg (-AB), De Vos 2282 (STE); 
Tygerberg near Prince Albert, Marloth 4452 (BOL, PRE). 

T. florentiae comprises small plants readily distinguished by their short 
stem, distichous, recurved leaves, pale papery bracts hidden by the leaf 
bases, and bright yellow flowers with the three lower perianth lobes showing 
an arrow-shaped red mark above a large triangular callus which has a 
crisped upper edge. 

The species has the habit, corm, and flower typical of Tritonia section 
Montbretia , but differs from all other Tritonia species in its inflorescence. 
On account of this L. Bolus placed it in a genus of its own, Monbretiopsis . 
The plant bears two to three, or rarely four, flowers each within its two 
bracts, terminal on a short peduncle which is produced in the axil of an up- 
per foliage leaf. This inflorescence can be interpreted as a very reduced 
branched spike, each branch with a single flower. This character is anal- 
ogous with that discussed by Goldblatt (1972) in Anomatheca, where one 
species, A. fistulosa has a branched inflorescence, each branch with a single 
terminal flower. The inflorescence of the present species differs from the lat- 
ter in its much shorter branches which are hidden by the leaf sheaths. 

Species Insufficiently Known 

Tritonia flava (Ait.) Ker-Gawl. in Kon. & Sims, Ann. Bot. 1: 228 (1805); 
Ait. Hort. Kew. ed. 2, 1: 92 (1810); non Ker-Gawl. 1823, nec Bak. 1892, 
1896. 

Gladiolus flavus Ait. Hort. Kew. 1: 65 (1789). Type: Hort. Kew. 1781 
(BM, holo-). 

Known only from its type in BM which is a very poor specimen compris- 
ing a scape with a single open flower and a few buds, a lateral branch with 
buds, two cauline leaves and basal leaf, and another loose basal leaf. 


African genus Tritonia Ker-Gawler: part 2 


413 


Scape ca. 150 mm long, flexuose with one branch. Basal leaves 130-140 
x 7-10 mm, lanceolate, acuminate, with a strong middle vein; cauline 
leaves smaller. Spike lax, with few flowers. Outer bract pale brown, 15-25 
mm long, narrowly lanceolate, acuminate, striate, with a stronger median 
vein and sphacelate papery margins; inner 15 mm long, bifid with 2 long acu- 
minate points. Flowers 40 mm long, yellow (e descr.), suberect. Perianth 
tube ca. 25 mm long, funnel-shaped; segments ca. 15 x 4-5 mm, the three 
lower segments with an unguiform perpendicular callus (e descr.). Anthers 5 
mm long. Style ca. 25 mm. 

Distribution unknown. 

It is not the same as T. flava Ker-Gawl. in Bot. Reg. 7 t. 747 (1823) (see 
under T. securigera ), nor is it the species designated as T. flava in South 
African herbaria (see T. karooica). 

T. fulva Dehnh. Cat. Horti Camald. ed. 2, 24 (1832) is, e descr., a Tritonia 
species, sect. Montbretia. Type not found. 

T. magniflora Dehnh. ibid, is, e descr., perhaps T. squalida (Ait.) Ker- 
Gawl. Type not found. 

T. pulchella Dehnh. ibid, is, e descr., perhaps T. squalida (Ait.) Ker-Gawl. 
Type not found. 

T. tripunctata Dehnh. ibid, is, e descr., a Tritonia species, section Montbre- 
tia. Type not found. 

T. undulata (N.L. Burm.) Bak. in J. Linn. Soc. 16: 163 (1877) partly as to 
name, excl. syn., & Handb. Irid. 191 (1892), & FI. Cap. 6: 121 (1896) as to 
name but not as to plant described, and excl. all syn. except I. undulata, 
Foster in Contr. Gray Herb. N.S. 114: 46 (1936) as to name but not as to 
plant; non Marloth (1915). 

Ixia undulata N.L. Burm. Prodr. FI. Cap. 1 (1768); non Salisb. (1796). 
Type: Africa australis, Caput bonae Spei, Herb. Burman s.n. (G, holo-). 

Tritonixia undulata (Bak.) Klatt in Abh. naturf. Ges. Halle 15. 357 
(1882) 

Tritonia undulata (N.L. Burm). N.E.Br. in Kew Bull. 1929. 137 (19-9). 

Scape ca. 120 mm long, slender, slightly flexuose. Leaves 3, ca. 60-100 x 
4-6 mm, acuminate, with crisped, finely ciliate, undulate margins and a 
strong middle vein. Spike lax, few-flowered. Bracts membranous, rown, 
5 mm long; outer ovate, subobtuse or shortly lacerate; inner bidentate 
Flowers 20-24 mm long, funnel-shaped, yellowish (fide Burman). Perianth 


414 


Journal of South African Botany 


tube 10-12 mm long, funnel-shaped; segments subequal, oblong-lanceolate 
(linear, fide Burman), obtuse, 10-11 x 3 mm, probably without calli. 
Anthers 2 mm long, curved, fairly stout for their length, reaching halfway up 
the perianth segments. Style ca. 16 mm long, reaching the anther tips. 

Distribution: Cape of Good Hope, without precise locality. 

Linnaeus the younger (1781) combined two species under Ixia crispa 
when he cited I. undulata Burm. as synonym. His description, however, re- 
fers to I. crispa only. Ker-Gawler [Curtis’s bot. Mag. t. 599 (1802)] also 
cited I. undulata as a synonym of I. crispa L.f. As Burman’s epithet was the 
earlier one, Baker (1877), on transferring the taxon to Tritonia , named it 
T. undulata (Burm.f.) and cited Linnaeus’s species as a synonym. The de- 
scription, provided in 1892 by Baker, refers, like Linnaeus’s description, to 
I. crispa only. 

The error was discovered by N. E. Brown (1929) when he examined the 
Iridaceae of Burman’s Prodromus. He thereupon described Burman’s plant 
as T. undulata (Burm.f.) N.E.Br. and gave I. crispa L.f. a new name, 
T. thunbergii, since the name T. crispa had already been used for another 
species of Tritonia. See T. thunbergii under Species Excluded. 

Foster (1936) realised that difficulties were involved in the acceptance of 
T. undulata (Burm.) N.E.Br. and indicated Burman’s species only as T. un- 
dulata (Burm.f.) Bak. “as to name but not as to plant”. 

Burman’s species is known only from its specimens in the Burman her- 
barium in Geneva. Superficially the plant looks like a species of Tritonia. Its 
anthers are, however, shorter and stouter than in any species of Tritonia , 
and it must perhaps be transferred to another genus, e.g. Ixia. 


Sections and Subgenera Excluded from Tritonia 

Section Crocosma (Planch.) Bak. in J. Linn. Soc. 16: 163 (1877) (orthogra- 
phic error) is Crocosmia Planch. 

Section Dichone Salisb. ex. Bak. in J. Linn. Soc. 16: 163 (1877) and sub- 
genus Dichone Salisb. ex Bak. Handb. Irid. 190 (1892) are Ixia subgenus 
Dichone Lewis in J1 S. Afr. Bot. 28: 159 (1962). 

Subgenus Stenobasis Bak. Handb. Irid. 196 (1892) is partly Tritoniopsis L. 
Bol. and partly Zygotritonia Mildbr. 

Section Stenobasis (Bak.) Diels in Pfl. Fam. ed. 2, 15a: 490 (1930) is Trito- 
niopsis L. Bol. 


African genus Tritonia Ker-Gawler: part 2 


415 


Species Excluded 

Montbretia abyssinica Hochst. in Flora 24: 1 (1841) [Tritonia abyssinica 
(Hochst.) Walpers, Ann. Bot. Syst. 3: 613 (1852-53)]. Type: Sckimper PL 
Abyss. 329: The isotype in G is a species of Geissorhiza (G. ?abyssinica R. 
Br. ex. Walp. p. 614) and not a Lapeyrousia as indicated by Baker (1877). 

T. acroloba Harms in Bot. Jb. 30: 278 (1902) is Radinosiphon lepto- 
stachys (Bak.) N.E.Br. in Trans. Roy. Soc. S. Afr. 20: 263 (1932). 

Montbretia aequimaculata Heynh. Norn. Bot. Hort. 2: 417 (1846) nom. 
nud. Identity not known. 

T. anigozanthiflora Sweet, Hort. Brit. ed. 1, 398 (1827) nom. nud. 
(Montbretia anigozanthaeflora (Sweet) Heynh. ibid., nom. nud.). Identity 
not known. 

T. aurea Pappe ex Hook, in Curtis’s bot. Mag. 73 t. 4335 (1847) is Cro- 
cosmia aurea (Pappe ex Hook.) Planch. FI, Serres Ser. 1, 7: 161 (1851-52). 

T. bongensis Pax in Bot. Jb. 15: 153 (1893) is Zygotritonia bongensis 
(Pax) Mildbr. in Bot. Jb. 58: 230 (1923). 

T. carnea Hort., Pasq. Cat. Orto Bot. Nap. 104 (1867), nom. nud. Iden- 
tity not known. 

T. cinnabarina Pax in Bot. Jb. 15: 152 (1893) is Crocosmia cinnabarina 
(Pax) De Vos, comb. nov. Type: Angola, Catala Canginga, Teucsz in Exped. 
A. v. Mechow 573 (B, holo-). 

T. concolor Sweet, Hort. Brit. ed. 1, 398: (1827) [ Montbretia concolor 
(Sweet) Voigt, Hort. Suburb. Calc. 611 (1845)] is Ixia paniculata Delaroche, 
Descr. PI. aliq. nov. 26, t. 1 (1766), fide Lewis (1962). 

T. cooperi Bak. Handb. Irid. 192 (1892) nom. illeg., non Klatt (1882) is 
Watsonia cooperi (Bak.) L. Bol. in J. Bot. Lond. 67:135 (1929). See also 
Lewis (1941). 

T. crocosmiflora Nich. Diet. Gard. 4: 94 (1888), a garden hybrid, is 
Crocosmia crocosmiflora (Nich.) N.E.Br. in Trans. Roy. Soc. S. Afr. 

264 (1932). 

T. fucata Lindl. in Bot. Reg. 24 t. 35 (1838) [Montbretia fucata (Lind!.) 


416 


Journal of South African Botany 


Voigt, Hort. Suburb. Calc. 611 (1845)] is Chasmanthe fucata (Lindl.) 
N.E.Br. in Trans. Roy Soc. S. Afr. 20: 274 (1932). 

Montbretia gallabatensis Schweinf. ex. Bak. in J. Linn. Soc. 16: 155 
(1877) is Lapeirousia abyssinica (Hochst.) Bak. ibid. 

T. graminifolia Bak. Handb. Irid. 195 (1892) is Anomatheca grandiflora 
Bak. in J. Bot. Lond. 14: 337 (1876), according to W. Marais’ determination 
of the type in K. 

T. latifolia (Delar.) N.E.Br. in Kew Bull. 1929: 135 (1929) is lxia latifolia 
Delar. Descr. pi. aliq. nov. 22 (1766), fide Lewis (1962). 

T. longiflora (Ker) Ker-Gawl. in Kon. & Sims, Ann. Bot. 1: 228 (1805), 
non N.E.Br. (1928) is lxia paniculata Delaroche Descr. pi. aliq. nov. 26 
(1766). 

T. macowanii Hort., cf. Gard. Chron. 2: 374, 407 (1877), an orthogra- 
phic error for Tritoma macowanii, is Kniphofia triangularis Kunth, Enum. 
PI. 4: 551 (1843). See Codd in Bothalia 9: 469. 

T. masonorum L. Bol. in Ann. Bol. Herb. 4: 43 (1926) is Crocosmia ma- 
sonorum (L. Bol.) N.E.Br. in Trans. Roy. Soc. S. Afr. 20: 264 (1932). 

T. mathewsiana L. Bol. in Ann. Bol. Herb. 3 : 76 (1923) is Crocosmia 
mathewsiana (L. Bol.) Goldbl. in J1 S. Afr. Bot. 37 : 423 (1971). 

T. mensensis Schweinf. in Bull. Herb. Boiss. 2, App. ii: 86 (1894) is a 
species of Gladiolus. Type: Eritrea: Gheleb in prov. Mensa, Schweinfurth 
1188 (K). 

T. media cf. Pritz. Ic. Ind. 1: 1124, an orthographic error for Tritoma 
media Ker-Gawl. which is Kniphofia sarmentosa (Andr.) Kunth, Enum. PI. 
4: 552 (1843). See Codd in Bothalia 9: 453. 

T. nervosa (Bak.) Klatt in Abh. naturf. Ges. Halle 15: 358 (1882) is Tri- 
toniopsis nervosa (Bak.) Lewis in J1 S. Afr. Bot. 25: 351 (1959). 

T. nervosa Bak. in Bull. Herb. Boiss. Ser. 2, 1: 864 (1901) nom. illeg., 
non Klatt (1882), is e. descr. a Tritoniopsis sp., cf. T. ramosa Eckl. ex Klatt 
var. unguiculata (Bak.) Lewis in J1 S. Afr. Bot 25: 329 (1959). 


African genus Tritonia Ker-Gawler: part 2 417 

T. odorata Lodd. Bot. Cab. 19 t. 1820 (1832), nom. nud. is Freesia co- 
rymbosa (Burm. f.) N.E.Br. in Kew Bull. 1929: 132 (1929). 

T. paniculata (Delaroche) Klatt in Abh. naturf. Ges. Halle 15 : 358 (1882) 
is Ixia paniculata Delaroche, Descr. pi. aliq. nov. 26 (1766), fide Lewis 
(1962). 

T. pauciflora (Bak.) Klatt in Abh. naturf. Ges Halle 15 : 358 (1882) 
[Montbretia pauciflora Bak. in J. Bot. Lond. 14 : 336 (1876)] is Gladiolus 
floribundus Jacq. Coll 4 : 162 (1792), f. Lewis e.a. (1972). 

T. pauciflora Bak. Handb. Irid. 193 (1892), non (Bak.) Klatt (1882), is 
Hesperantha pauciflora (Bak.) Lewis in Flow. PI. S. Afr. 18 t. 682 (1938). 

T. pottsii (Bak.) Bak. in Curtis’s bot. Mag. 109 t. 6722 (1883) is Crocos- 
mia pottsii (Bak.) N.E.Br. in Trans. Roy. Soc. S. Afr. 20 : 264 (1932). 

T. quinquenervata Foster in Contr. Gray Herb. 114 : 46 (1936) is 
Watsonia cooperi (Bak.) L. Bol. in J. Bot. Lond. 67: 135 (1929). See also 
Lewis (1941). 

T. reflexa Klatt ined. in herb, is Ixia scillaris L. 

T. refracta (Jacq). Ker-Gawl. in Kon. & Sims, Ann. Bot. 1: 228 (1805) is 
Freesia refracta (Jacq). Klatt in Linnaea 34: 673 (1865-66), f. Goldblatt 
(1982). 

T. riparia Cordemoy, FI. Pile Reunion 161 (1895) is possibly a species of 
Dietes. The isotype in P has fruits only. 

T. rocheana Sweet, Hort. Brit. 398 (1827) [M. rocheana Heynh. Nom. 
Bot. Hort. 2: 418 (1846)] is Ixia bellendenii Foster in Contr. Gray Herb. 
114 : 47 (1936). 

T. rochensis Ker-Gawl. in Curtis’s bot. Mag. 37 t. 1503 (1812) is Ixia bel- 
lendenii Foster ibid. cf. Lewis (1962). 

T. schimperi Aschers. & Klatt in Linnaea 34: 697 (1865-66) is 
Lapeirousia schimperi (Aschers. & Klatt) M. Redh. in Kew Bull. 34. 307. 
Type: Schimper PL Abys. 2304 (G, iso-). 


T. schlechteri Bak. in Bull. Herb. Boiss. Ser. 2, 4 . 1005 (1904) is a 


418 Journal of South African Botany 

species of Ixia , cf. I. orientalis L. Bol. Type: Cape, Swarteberg, Schlechter 
5609 (Z, holo-). 

T. scillaris (L.) Bak. in J. Linn. Soc. 16: 163 (1877) is Ixia scillaris L. Sp. 
PI. ed. 2, 1: 52 (1762), fide Lewis (1962). 

T. striata (Jacq.) Ker-Gawl. in Kon. & Sims, Ann. Bot. 1: 228 (1805), 
non Klatt 1895, [Montbretia striata (Jacq.) Voigt, Hort. Suburb. Calc. 610 
(1845)] is Babiana striata (Jacq.) Lewis in J1 S. Afr. Bot. 7: 55 (1941). 

T. templemannii Bak. Handb. Irid. 193 (1892) is Pillansia templemannii 
(Bak.) L. Bol. in Ann. Bol. Herb. 1: 20 (1915). 

T. tenuiflora (Vahl) Ker-Gawl. in Curtis’s bot. Mag. 31 sub t. 1275 
(1810) [Montbretia tenuiflora (Ker-Gawl.) Voigt, Hort. suburb. Calc. 611 
(1845)] is Ixia paniculata Delaroche, Descr. pi. aliq. nov. 26, t. 1 (1766), fide 
G. J. Lewis (1962). 

T. teretifolia Bak. Handb. Irid. 194 (1892) is Gladiolus permeabilis 
Delar. Descr. pi. aliq. nov. 27, t. 2 (1766) ssp. wilsonii (Bak.) Lewis e.a. 
(1972). 

T. thunbergii N.E.Br. in Kew Bull. 1929: 137 (1929) is Ixia erubescens 
Goldbl. in J1 S. Afr. Bot. 37: 233 (1971). 

T. tigrina Pax in Bot. Jb. 15: 152 (1893) is possibly a species of Schizosty- 
lis. From Angola. Type in B. 

T. trinervata Bak. Handb. Irid. 191 (1892) is Ixia trinervata (Bak.) Lewis 
inJIS. Afr. Bot. 28: 169 (1962). 

T. tristis Dehnh. Cat. horti Camald. ed. 2, 24 (1832) is, e descr., not Tri- 
tonia. Type not found. 

T. undulata sensu Bak. in J. Linn. Soc. 16: 163 (1877) in part, non 
N.E.Br., is Ixia erubescens Goldbl. in J1 S. Afr. Bot. 37: 233 (1971). 

T. unguiculata Bak. Handb. Irid. 196 (1892) is Tritoniopsis ramosa Eckl. 
ex Klatt var. unguiculata (Bak,) Lewis in J1 S. Afr. Bot. 25: 329 (1959). 

T. ventricosa Pasq. Cat. Orto Bot. Nap. 104 (1867) nom. nud. Identity 
not known. 


419 


African genus Tritonia Ker-Gawler: part 2 

T. ventricosa Bak. Handb. Irid. 193 (1892) is Gladiolus brevitubus Lewis 
in J1 S. Afr. Bot. Suppl. 10: 171 (1972). a nom. nov. for T. ventricosa Bak. 

T. viridis (Ait.) Ker-Gawl. in Kon. & Sims, Ann. Bot. 1: 231 (1805) 
[Montbretia viridis (Ker-Gawl.). Voigt, Hort. suburb. Calc. 610 (1845)] is 
Anomatheca viridis (Ait.) Goldbl. in J1 S. Afr. Bot. 37: 443 (1971). 

T. watsonioides Bak. Handb. Irid. 195 (1892) is Watsonia watsonioides 
(Bak.) Oberm. in Bothalia 8: 115 (1962). 

T. wilsonii Bak. in Gard. Chron. 26: 38 (1886) is Gladiolus permeabilis 
Delar. ssp. wilsonii (Bak.) Lewis in J1 S. Afr. Bot. Suppl. 10: 133 (1972). 

T. xanthospila Ker-Gawl. ex Spreng. Syst. Veg. 1: 154 (1825) [Montbre- 
tia xanthospila Heynh. Nom. Bot. Hort 2: 418 (1846)] is Freesia xanthospila 
Klatt in Linnaea 34: 673 (1865-66), a species based on cultivated material, f. 
Goldblatt in J1 S. Afr. Bot. 48: 89 (1982). 

Addendum: Note on a Species Already Revised in Part 1 

Tritonia tugwelliae L. Bol. 

The following more detailed description is given of the flowers from a 
collection, De Vos 2463, found in 1981, east-north-east of Laingsburg along 
the N1 national road (Ladismith 3321-AA), that flowered in August 1982. 

Perianth tube with red lines inside; segments (except the median posti- 
cous) with red lines in the throat; median anticous also with a yellow zone in 
the throat above which a small red spot occurs. Stamens and style curved 
against the posticous segment; anthers at first violet, pollen violet or mauve. 
Stigmas small, terminal. 

The newly opened flowers have the median posticous perianth segment 
porrect and the other segments spreading or recurved. In older flowers, the 
three anticous segments fold upwards over one another, partially closing the 
throat of the flower; the three posticous segments are strongly reflexed, also 
folding over one another at the posterior side of the flower. Fig. 30. 

Acknowledgements 

My grateful thanks are due to the directors and curators of the herbaria 
mentioned in this work, from whom specimens of Tritonia, including type 
specimens, were obtained for examination; to the many people who helped 
in collecting fresh material from their natural habitats; and lastly, to the 
head of the Department of Botany of the University of Stellenbosch for fa- 
cilities to work in the Department after my retirement. 


420 


Journal of South African Botany 



Fig. 30 

Tritonia tugwelliae (De Vos 2463): a, plant; b, outer (left) and inner (right) bract; c, 
newly opened flower; d, older flower; e, capsule; an, median anticous perianth seg- 
ment; p, posticous segment. 


African genus Tritonia Ker-Gawler: part 2 


421 


References 

Aiton, W., 1789. Hortus Kewensis 1. London: Nicol. 

Aiton, W., 1810. Hortus Kewensis ed. 2, 1. London: Longman e.a. 

Baker, G. J., 1875. Montbretia laxifolia. Trans. Linn. Soc. 29: 155. 

Baker, J. G., 1876. New species of Ixiae. J. Bot. Lond. 14: 237. 

Baker, J. G., 1877. Systema Iridacearum. J. Linn. Soc. 16: 161. 

Baker, J. G., 1892. Handbook of the Irideae. London: Bell & Sons. 

Baker, J. G., 1896. Irideae. In: Thiselton-Dyer, Flora Capensis 6. London: Reeve. 
Batten, A. and Bokelmann, H., 1966. Wild flowers of the Eastern Cape Province. 
Cape Town: Books of Africa. 

Bentham, G. and Hooker, J. D., 1883. Genera Plantarum 3. London: Reeve. 
Brown, N. E., 1928. The South African Iridaceae of Thunberg’s herbarium. J. Linn. 
Soc. Bot. 48: 15. 

Brown, N. E., 1929. The Iridaceae of Burman’s Florae Capensis prodromus. Kew 
Bull. 1929: 129. 

Burman, N. L., 1768. Prodromus Florae Capensis. Leiden: Haak. 

De Candolle, A. P., 1803. Montbretia. Bull. Soc. Philom. 3: 151. 

Delaroche, D., 1766. Descriptiones plantarum aliquot novarum. Leiden: Verbeek. 
De Vos, M. P., 1982a. Die bou en ontwikkeling van die unifasiale blaar van Tritonia 
en verwante genera. Jl S. Afr. Bot. 48: 23. 

De Vos, M. P., 1982b. The African genus Tritonia Ker-Gawler: Part I. Jl S. Afr. 
Bot. 48: 105. 

Ecklon, C. F., 1827. Topographisches Verzeichnis der Pflanzenversammlung von 
C. F. Ecklon. Esslingen: Reise-Verein. 

Foster, R. C., 1936. Notes on nomenclature in Iridaceae. Contr. Gray Herb. Harv. 
N.S. 114: 44. 

Goldblatt, P., 1972. A revision of the genera of Lapeirousia and Anomatheca in the 
winter rainfall regions of South Africa. Contr. Bolus Herb. 4: 78. 
Goldblatt, P., 1976. Iridaceae. In: Dyer, R. A., The genera of Southern African 
flowering plants 2. Pretoria: Dept. Agric. Techn. Services. 

Goldblatt, P., 1982. Systematics of Freesia Klatt. Jl S. Afr. Bot. 48: 39. 

Goldblatt, P. and Barnard, T. T., 1970. The Iridaceae of Daniel De la Roche. Jl 
S. Afr. Bot. 36: 291. 

Houttuyn, M., 1780. Natuurlijke Historie 2, 12. Amsterdam: F. Houttuyn. 

Jacquin, N. L., 1789. leones plantarum rariorum 2. Vienna: Wappler. 

Jacquin, N. L., 1796. Collectanea 5. Vienna: Wappler. 

Ker-Gawler, J. B., 1800. Gladiolus lineatus. In: Curtis’s bot. Mag. t. 487. 
Ker-Gawler, J. B.. 1803. Tritonia capensis. In: Curtis’s bot. Mag. t. 618. 
Ker-Gawler, J. B., 1805. Ensatarum ordo: Tritonia. In: Kon. & Sims, Ann. Bot. 1 : 
227. 

Ker-Gawler, J. B., 1810. Tritonia viridis. In: Curtis’s bot. Mag. sub t. 1275. 
Ker-Gawler, J. B., 1813. Tritonia capensis. var. minor. In: Curtis’s bot. Mag. 
t. 1531. 

Klatt, F. W., 1863. Revisio Iridearum. Linnaea 32: 752. 

Klatt, F. W., 1882. Erganzungen u. Berichtigungen zu Baker’s Systema Iridacea- 
rum. Abh. naturf. Ges. Halle 15: 355. 


422 


Journal of South African Botany 


Klatt, F. W., 1895. Irideae. In: Durand & Schinz, Conspectus florae Africae 5: 203. 
Letty, C., 1962. Veldblomme van Transvaal. Pretoria: Dept. Landbou. 

Lewis, G. J., 1941. Iridaceae. New genera and miscellaneous notes. Jl S. Afr. Bot. 7: 
19. 

Lewis, G. J., 1959. The genus Babiana. Jl S. Afr. Bot. Suppl. Vol. 3. 

Lewis G. J., 1962. The genus Ixia. Jl S. Afr. Bot. 28: 45. 

Lewis, G. J., Obermeyer, A. A. and Barnard, T. T., 1972. Gladiolus. Jl S. Afr. 
Bot. Suppl. Vol. 10. 

Linnaeus, C.f., 1781. Supplementum plantarum. Brunsviga: Orphanotrophei. 
Nordenstam, B., 1972. Types of Ecklon’s Topographisches Verzeichnis in the 
Swedish Museum of Natural History in Stockholm. Jl S. Afr. Bot. 38: 277. 
Pearse R., 1978. Mountain splendour: Wild flowers of the Drakensberg. Cape Town: 
Timmins. 

Phillips, E. P., 1951. The genera of South African flowering plants. Mem. bot. Surv. 
S. Afr. 25. 

Redout^, P., 1804. Les Liliacees 1. Paris: Didot. 

Redout£, P., 1813. Les Liliacees 7. Paris: Didot. 

Salisbury, R. A., 1796. Prodromus stirpium in horto ad Chapel Allerton vigentium. 
London. 

Schlechter, R. 1908. Tritonia flavida. Bot. Jb. 40: 91. 

Sealy, R. J., 1939-40. Tritonia flavida. In: Curtis's bot. Mag. 162 sub t. 9592. 
Thunberg, C. P., 1784. Dissertatio de Gladiolo. Upsala: Edman. 

Vahl, M., 1805. Enumeratio plantarum 2. Copenhagen: Moller. 

Verdoorn, I. C., 1959. Tritonia nelsonii. Flow. PI. Afr. 33 PI. 1315. 

Voigt, J. O., 1845. Hortus suburbanus Calcuttensis. Calcutta: Bishops College Press. 
Willdenow, C. L., 1797. C. Linne, Species plantarum , ed. 4, 1. Berlin: Nuak. 


J1 S. Afr. Bot. 49 (4): 423-444 (1983) 


SIX MORE NEW SPECIES OF LACHENAL1A (LILIACEAE) 


W. F. Barker 


(Compton Herbarium, Kirstenbosch, Private Bag X7, Claremont 7735 
R.S.A. 

Bolus Herbarium, University of Cape Town, Private Bag, Rondebosch 7700, 
R.S.A.) 


Abstract 

Six new species of Lachenalia are described, and the characters of the ripe seeds 
are used as additional diagnostic features in the genus. 


UlTTREKSEL 

’N VERDERE SES NUWE LAC H ENA LI A-SOORTE (LILIACEAE) 

Ses nuwe Lachenalia - soorte word beskryf en die eienskappe van die ryp sade 
word as aanvullende onderskeidende kenmerke in die genus voorgestel. 

Key words: Lachenalia, sp. nov., Liliaceae, South Africa, seed, diagnostic charac- 
ters. 


Introduction 

The six new species described all have seeds with terminal ridged arils of 
various lengths. Three of the species, L. zebrina, L. nordenstamii and 
L. whitehillensis, have a number of features in common, such as the single 
banded leaf, campanulate flowers with exserted stamens, and in the case of 
the two former the seeds are the largest in the genus, with medium-length 
terminal ridged arils and extruded micropyles. In L. ameliae the seed is very 
small, with a medium ridged aril, the flowers are urceolate and the leaves of- 
ten have an integument of hairs on the upper surface. L. giessii and L. mon- 
iliformis have seeds which are very small with an almost obsolete terminal 
ridged aril. L. giessii is confined to the south western area of South West 
Africa/Namibia, while L. moniliformis is unique in having many filiform 
leaves with a beaded appearance, and is only known from the type locality 
near Worcester, where it was discovered as recently as 1978. 


Accepted for publication 17th June, 1983 


423 


424 


Journal of South African Botany 



Fig. 1. 

Lachenalia zebrina: 1. Outer perianth segment and stamen; 2. Inner perianth seg- 
ment; 3. Gynaecium; 4. Ripe capsule; 5. Capsule from above; 6. Seed ( Compton 

3536). 

Lachenalia zebrina Barker, sp. nov.; folio solitario, conduplicato vel ca- 
naliculate, falcato vel lanceolate, supra glauco, subter fasciato marronino, 
basi amplecti atropurpurescenti, distinguitur. 

Inflorescentia racemosa; pedicelli longi vel breves, patentes. Flores cer- 
nui, curti campanulati, basim truncatam; segmenta externa ovalia, eburnea 
suffusa viridia vel brunnea, apices recurvatos; segmenta interiora leviter 
longiora vel segmenta externa; stamina manifeste exerta. Capsula grandis, 
inflata, oboviodea. Semen magnum, elongatum, plerumque curvatum; aril- 
lus porcatus, terminalis; micropyle manifeste exertus, porcatus, arillum 
aequantes. 

Type Material: Cape Province — 3320 (Montagu): Karoo Garden, Whitehill 
(-BA), 17/9/1945, Compton 17392 (NBG, holo., BOL, iso.). 


Six new species of Lachenalia (Liliaceae) 


425 


Plant up to 300 mm, usually less. Bulb globose to obovoid 10-30 mm 
diam.; outer tunics often spongy, light brown. Leaf one, blade 60-220 mm 
long, 20-60 mm broad, falcate to lanceolate, conduplicate or channelled, 
glaucous green above, banded with maroon below, margin often undulate, 
clasping base, silvery-white with maroon bands above shading to magenta at 
base. Inflorescence racemose, 60-105 mm long, 20-60 mm diam., few- to 
many-flowered, lax to dense, pedicels 2-20 mm long, often shorter toward 
the base, spreading, flowers cernuous, bracts minute, membranous, narrow 
lanceolate. Flowers small, cream tinged with green or brown, 4-6 mm long, 
tube a flat disc 2 mm diam.; outer segments 4-6 mm long, oblong with apex 
recurved; inner segments slightly longer, narrower with recurved apex; sta- 
mens cream, well exserted, up to 8 mm long; ovary ovoid 1-2 mm diam., 
green; style 4-8 mm long. Capsule membranous, globose to obovoid up to 



Fig. 2. 

Lachenalia zebrina: forma densiflora ( Thomas 


sub NGB 105714). 


426 


Journal of South African Botany 


10 mm diam., winged; seeds large, elongate, curved, aril ridged, terminal, 
medium length, micropyle extruded, ridged, as long as aril. 

Diagnostic Characters 

Lachenalia zebrina is characterised by its single, usually falcate, con- 
duplicate or channelled, banded leaf which is glaucous green above; by its 
racemose inflorescence; and flowers which are truncate at the base; with the 
tube reduced to a small flat disc; by its large, inflated capsule, which is one 



Lachenalia zebrina: 1. Seed side view; 2. Seed from above; 3. Seed from below 

{Compton 21939). 

of the largest in the genus; and its comparatively large curved seed, with a 
medium length ridged aril and an extruded ridged micropyle of equal length. 

Lachenalia zebrina Barker, sp. nov. forma densiflora Barker forma nov.; 
a forma typica, inflorescentia densa, pedicellis brevibus et floribus patenti- 
bus, differt. 

Type Material: Cape Province — 3119 (Calvinia): 23 mis N of Downes on Klipwerf 
road (-BD), 2/10/1974, Thomas s.n. sub NBG 105714 (NBG, holo.). 

It differs from the typical form in its more numerous flowers with short 
pedicels, forming a dense narrow inflorescence. 

It is not possible to separate forma densiflora geographically on its distri- 
bution pattern, but it is such a variable species that it is considered useful to 
designate the two extremes of the graduated sequence. 

Specimens of Lachenalia zebrina have previously been included in the 
species Lachenalia anguinea Sweet, and in 1930 it was mistakenly illustrated 
under that name, in the Journal of The Botanical Society XVI Plate II pp. 
10-12. However, recently, when a study was made of the capsules and ripe 
seeds of the two taxa, it was found that the plants which occur in the coastal 


Six new species of Lachenalia (Liliaceae) 


427 


areas of the western Cape, had a small obovoid capsule containing very 
small, ovoid seeds with a minute ridged, terminal aril, and the flowers had a 
distinct tube which narrowed towards the base. The capsules of the plants 
occurring in the Karoo areas were very large, inflated and globose, and the 
seeds were very much larger, elongated, curved, with a ridged terminal aril 
of medium length, and an exstruded, ridged micropyle of about the same 
length as the aril, while the perianth was truncate at the base. 

These taxa are now considered to represent two distinct species, separ- 
ated geographically. The plants from the sandy coastal areas of the western 
Cape, extending from Piketberg in the south to the Richtersveld in the 
north, agree with Sweet’s illustration in Sweet’s British Flower Garden I 
t. 179 (1838), which represents the iconotype of L. anguinea Sweet. The 
laid-out flower is depicted with a distinct tube. 

The plants, which are widely distributed in the Karoo from Matjiesfon- 
tein in the south, to the Knersvlakte in the Vanrhynsdorp district in the 
north, and eastwards to the Carnarvon, Williston, and Sutherland districts, 
are extremely variable in the width of the leaves and the shape of the inflor- 
escence, but are all grouped under the new name Lachenalia zebrina. 

In the typical form the inflorescence is very lax with long pedicels to the 
flowers, which are cernuous; in forma densiflora the flowers are much more 
numerous and congested, with comparatively short pedicels and the flowers 
are spreading, producing a dense, narrow elongated head, which gives the 
plant a distinctive appearance. However, intermediate variations occur 
which connect the two extremes. 

The specific name was suggested by the resemblance of the striped, fal- 
cate leaf to the hind leg of a zebra. 

L. zebrina Barker forma zebrina 


Specimens Examined 

CAPE PROVINCE— 3022 (Carnarvon): Carnarvon (-CC), Pillans s.n. Hort. 
NBG 3740114 (BOL). , _ „ „ 

— 3118 (Vanrhynsdorp): Farm Klipdrift, Knersvlakte (-B), 22/9/1970, Hall 362 

—3119 (Calvinia): Brandkop, NE of Nieuwoudtville (-AA), 1319/1961, barker^ 9484 
(NBG); Near Kokerboomkop 16 mis S of Loeriesfontein (-AB), 27/9/1952, Lewis 
2517 sub SAM 61824 (BOL, K, SAM); Between Loeriesfontein and Nieuwoudtville, 
27/9/1952, Johnson 604 (NBG); Gannabos junction with Loeriesfontein road, 
2/10/1974, Thomas s.n. sub NBG 105717 (NBG); NE of Nieuwoudtville on Kleindo- 
ring road (-AD), 28/9/1970, Barker 10765 (NBG); 30 mL of Calvinia (-BC), 
29/10/1970, Stayner s.n. sub NBG 93033 (NBG); Calvinia (-BD), S e pt l 914 M^. 
pherson s.n. sub Hort. NBG 3693/14 (BOL); Calvinia, 1/10/1940 ^her 85. 3 (NBG). 
—3120 (Williston): Farm Langfontein SE of Calvinia (-CA) 25/8/1973 Hall 42 
(NBG); Visagiesrus near Calvinia (-DA), 8/1956, Acocks 18894 (P ), 

Williston (-D ), 24/9/1954, Acocks 17710 (PRE). 


428 


Journal of South African Botany 


— 3121 (Fraserburg): Between Fraserburg and Carnarvon (-D), 20/9/1938, Hafstrom 
and Acocks 222 (PRE); 42 km W of Loxton (-DC) 29/9/1974, Nordenstam and Lund- 
gren 2090 (S). 

— 3219 (Wuppertal): Elands Vlei, Lower Tanqua (-BC), 8/9/1921, Marloth s.n. sub 
10461 (PRE); Gemsbokfontein, Ceres Karoo (-DB), 15/9/1973, Hanekom 2149 
(PRE). 

— 3220 (Sutherland): Foot of Gannagas Pass (-AA), 29/9/1968, Stay ner s.n. sub NBG 
93587 (NBG); Top of Gannagas Pass, 29/9/1968, Stayner s.n. sub NBG 93586 
(NBG); Hout Hoek (-CA), 13/9/1971, Hanekom 1574 (PRE). 

— 3320 (Montagu): Patats Rivier, Ceres, Karoo (-AB), 2/10/1954, Esterhuysen 23535 
(NBG, BOL, K, PRE); Whitehill Ridge (-BA), 26/9/1926, Compton 3137 (BOL); 
Karoo Garden, Whitehill, 24/9/1929, Compton 3536 (BOL, NBG), 12/10/1930, 
Compton s.n. sub BOL 21939 (BOL, NBG), 22/9/1941, Barker 1093 (NBG), 
23/10/1943, Compton 15240 (NBG), 17/9/1945, Compton 17392 (BOL, iso.; NBG, 
holo.); Whitehill, 20/9/1943, Compton 14843 (NBG); Whitehill, 18/10/1928, J. Gillett 
1712 (STE); 2 mis NE of Matjiesfontein, 3/10/1974 (young fruit), 16/8/1976 (flowers), 
S. Olivier s.n. sub NBG 105718 (NBG); Skeiding 7 mis beyond Matjiesfontein, 
10/10/1954 (fruiting), Isaac s.n. sub NBG 71842 (NBG); 5 km E of Matjiesfontein, 
28/9/1974, Nordenstam and Lundgren 2080 (S). 


L. zebrina Barker forma densiflora Barker 


Specimens Examined 

CAPE PROVINCE — 3119 (Calvinia): Near Soetwater at crossroads to Botter- 
kloof (-AD), 1/10/1973, Barker 10891 (NBG); 23 mis N of Downes on Klipwerf 
Road (-BD), 2/10/1974, Thomas s.n. sub NBG 105714 (NBG, holo.); Road from 
Farm Kleinbrak to Middlepos (-DD), 28/9/1971, Hiemstra s.n. sub Hort. NBG 
903/1971 (NBG). 

— 3120 (Williston): 20 mis E of Calvinia on Brandvlei road (-AC), 27/10/1970, 
Thomas s. n. sub NBG 93871 (NBG). 

Lachenalia nordenstamii Barker, sp. nov.; habitu nano, bulbo multi-tuni- 
cato, strato extimo spongioso et strato intimo fibris membranis fissis, distin- 
guitur. 

Folium solitarium, lanceolatum-falcatum, canaliculatum; supra glaucum, 
subter fasciato vel maculato marronino. Pedunculus brevis valde gracilis. 
Inflorescentia racemosa, paucifloribus. Flores pusilli, cernui, late campanula- 
ti; tubus brevis; segmenta aequalia, bruneola, stria media marronina; apices 
recurvatos. Stamina valde exerta; filamenta crassa, marronina patentia. Cap- 
sula grandis, obovata vel cordata, alata. Semen elongatum vel rectum, inter- 
dum, curvatum; arillus porcatus micropyle exertus, porcatus, arillum ae- 
quans. 


Type Material: Cape Province — 2816 (Oranjemund): Numees Mts. (N of Hells- 
kloof) Richtersveld, Namaqualand (-BD), bulbs collected 3/11/1962, flowered in cul- 
tivation 31/5/1963, B. Nordenstam 1739 (NBG, holo. of flowering plant); Mountain 
between Numees and Hellskloof, Richtersveld, Namaqualand, 4/11/1962, B. 
Nordenstam 1762 (NBG, holo. of capsules and seeds). 


Six new species of Lachenalia ( Liliaceae ) 


429 


Plant dwarf 50-120 mm high. Bulb up to 20 mm diam., globose to ovate, 
covered with many layers of spongy tunics, produced into a neck up to 40 
mm long, composed of many tunics, the inner ones split into narrow mem- 
branous fibres. Leaf one, 20-40 mm long, lanceolate, falcate, canaliculate, 
glaucous green, banded on the clasping base and lower surface with maroon 
bands, sometimes merging into spots towards the apex. Peduncle very slen- 
der, tinged with maroon, up to 25 mm long above the neck. Inflorescence 
racemose, few-flowered; bracts minute, narrow-lanceolate, membranous; 
pedicels short, 2-3 mm long, spreading to decurved. Flowers small, cernu- 
ous, widely campanulate, tube short, segments almost equal in length, 4-5 
mm long, brownish with a central maroon stripe, apices recurved; stamens 
well exserted, up to 10 mm long, filaments comparatively stout, maroon, 
widely spread; style slender finally longer than the stamens. Capsule broadly 
obovate to obcordate, winged, up to 12 mm diam., 5-6 mm long, purplish; 
seeds large 3-3,5 mm long, 1-1,5 mm diam., usually straight, occasionally 
curved, aril ridged, medium, terminal, micropyle ridged, exstruded, as long 
as the aril. 


Diagnostic Characters 

Lachenalia nordenstamii is distinguished by its dwarf habit, the long 
fibrous neck to the bulb, its single lanceolate, falcate canaliculate, banded 
leaf, its few-flowered racemose inflorescence, in which the short pedicels are 
spreading to decurved, its cernuous short widely campanulate flowers, with 
the stamens twice the length of the perianth, and the stout maroon filaments 
spreading widely. The seeds are some of the largest in the genus, distinctly 
longer than broad, usually straight, with a ridged aril of medium length, and 
an exstruded micropyle of equal length. 

Two collections of bulbs of this very dwarf species were made by Dr. 
Bertil Nordenstam in the Richtersveld in November 1962. One of the collec- 
tions was in fruit, and the capsules were preserved, and now represent the 
type specimens of the fruiting stage of the species. Bulbs of the second col- 
lection were grown in a pot at the Compton Herbarium, and one of these 
flowered on 31 May 1963. This flowering specimen was preserved and now 
represents the type of the species. Fortunately photographs were taken to 
supplement the material by Professor H. B. Rycroft, in both black and 
white, and colour, and these have aided considerably in describing the 
species. Two of these photographs are represented here. 

In a recent consignment of Lachenalia specimens received on loan rom 
Munich for identification, two good sheets of this species, one of flowering 
specimens and the other in the fruiting stage, were included. They were co 
lected in 1977 in the south western corner of South West Africa/Namibia, 


430 


Journal of South African Botany 


a 


^ ' ^3* 3 9 . /y -ivcsv^c 'Ajiyn/iTustrvu, 

/Vvvn^t ts\ A'£i? viasrfUr 

( H trj H^CU J , ■^-•AUw 

i -V»-t4. t-V-e-v-i/'ce-, 

^ 4" A*y—jif~~L c*£ 


-^€Ay 


> , . . , teJ 3./'. ma 

•j , J*<C 3i~~% Iftoi 





*#•#*£ 

lr VP 10 

|,TF»^rTTr T rni[m^i^ii|MM(iiiM 1 ^ l | Mr, l"’ l l'' l, t ,, "\" ,Ml, "l ? l,n 


Fig. 4. a, b, c. 

Lachenalia nordenstamii: a. Whole plant; b. Three plants ( Nordenstam 1739)\ 
c. Capsules, seeds and old perianths ( Nordenstam 1762). 


431 


Sa new species of Lachenalia (Liliaceae) 



Fig. 5. 

Lachenalia nordenstamii: 1. Seed side view; 2. Seed from above; 3. Seed from below 

(Nordenstam 1762). 

across the Orange River, north of the Richtersveld. They had been iden-^ 
tified as Lachenalia buchubergensis Dinter, which has a similar leaf, and oc- 
curs in the same area; however in L. buchubergensis the flower is very dis- 
tinct, it has a narrow cylindrical, sessile flower, with the stamens just 
emerging from the narrow mouth of the flower. The Zurich specimens are 
the first of L. nordenstamii to be recorded from SWA/Namibia, and they in- 
dicate that its distribution should be further investigated. 

Lachenalia nordenstamii is closely allied to Lachenalia zebrina Barker 
sp. nov., but the latter is a much taller plant, without the fibrous neck to the 
bulb, the inflorescence is many-flowered and the flowers are narrowly cam- 
panulate, with slender declinate filaments. The seeds in both species are of a 
similar pattern, but in L. zebrina they are shorter and usually curved. 

The distribution of the two species does not appear to overlap, as L. 
nordenstamii , as far as is known, occurs in the northern area of the Richters- 
veld and the southern part of SWA/Namibia, while L. zebrina is wide- 
spread, occurring in many parts of the Karoo. The most north westerly rec- 
ord of it available at present, is a collection made by Mr. H. Hall on the 
farm Klipdrift on the Knersvlakte, in the Vanrhynsdorp district. 

The species is named in honour of Dr. Bertil Nordenstam who has done 
so much valuable work on the South African Flora, including some mem- 
bers of the Liliaceae. 

Specimens Examined 

SOUTH WEST AFRICA/NAMIBIA— 2716 (Witputs): Aurus Mountain, South 
West Africa (-CA), 31/7/1977 (flowering), M. Muller 744 (M); Rosh Pinah und 


432 


Journal of South African Botany 


Umgebung, Distrikt Liideritz-Sud, South West Africa (-DD), 29/9/1977 (fruiting), 
H. Merxmiiller and W. Giess 32356 (M,W?). 

CAPE PROVINCE — 2816 (Oranjemund): Numees Mts (N of Hellskloof) Rich- 
tersveld, Namaqualand (-BD), bulbs collected 3/11/1962, flowered in cultivation 
31/5/1963, B. Nordenstam 1739 (NBG, holo. of flowering plant); Mountain between 
Numees and Hellskloof, Richtersveld, Namaqualand 4/1 1/1962, B. Nordenstam 1762 
(NBG, holo. of capsules and seeds). 

Lachenalia whitehillensis Barker, sp. nov.; folio solitario, conduplicato, 
angusto-lanceolato vel lineari, supra veneto subter maculato vel fasciato 
marronino basi amplecti atropurpurescenti, distinguitur. 

Inflorescentia racemosa, angusta; pedicelli curti; flores cernui vel patent- 
es. Flores margaritacei vel opalescentes, campanulati, basim truncatam; seg- 
menta interiora longiora vel segmenta externa; segmenta laterale imbricata. 
Stamina manifeste exerta. Semen oviodeum; arillus porcatus terminalis. 

Type Material: Cape Province — 3320 (Montagu): Whitehill Laingsburg Division 
(-BA), 20/9/1943, Compton 14845 (NBG, holo.). 

Plant up to 360 mm high. Bulb globose, up to 20 mm diam.; outer tunics 
membranous, brown. Leaf 1, blade narrow-lanceolate, conduplicate, up to 
120 mm long, 15 mm wide blue-green, spotted or banded with maroon on 
lower surface, clasping base up to 80 mm long, spotted or banded with mar- 
oon shading to magenta toward the base. Peduncle slender, up to 140 mm 
long, banded or spotted with maroon. Inflorescence racemose, few- to many- 
flowered up to 160 mm long, 20-30 mm diam.; rachis spotted with maroon; 
flowers campanulate, truncate at base, spreading or slightly cernuous, 
scented; pedicels up to 3 mm long. Perianth up to 7 mm long, 4 mm diam. at 
base; tube 1 mm long; outer segments ovate, up to 4 mm long, pale blue at 
base, cream above with pale red-brown gibbosities; inner segments 6 mm 
long, obovate-obtuse, cream with pale blue keel and a brownish spot near 
the apex; two lateral segments overlapping and concave; stamens exserted 
up to 10 mm long; ovary ovate, pale green; style up to 10 mm long. Capsule 
membranous, obovate, 7 mm long; seed ovoid, aril ridged, terminal, me- 
dium length. 

Diagnostic Characters 

Lachenalia whitehillensis is characterised by its single linear to narrow- 
lanceolate, conduplicate leaf which is spotted or banded on the lower sur- 
face and on the clasping base. Inflorescence racemose, narrow, short ped- 
icels. The pale campanulate flowers are truncate at the base, the inner seg- 
ments longer than the outer, and the two lateral inner ones overlapping and 
concave. The stamens are well exserted and the ovoid seeds have a ridged 
terminal aril. 


Six new species of Lachenalia (Liliaceae) 


433 



Fig. 6. 

Lachenalia whitehillensis: 1. Flower side view; 2. Flower from below; 3. Flower front 
view; 4. Upper outer perianth segment; 5. Lateral outer perianth segment; 6. Lateral 
inner perianth segment and stamen; 7. Lower inner perianth segment; 8. Gynaecium 

( Compton 14845). 


Lachenalia whitehillensis superficially resembles L. zebrina which occurs 
in the same area in the type locality, but it differs from it in the shape and 
colouring of the flowers, in its narrow inflorescence, and markedly in the 
shape of the seed which in L. zebrina is elongate and one of the largest in 
the genus. 

The species has been named after the original site of the Karoo Garden 
near Whitehill Station, where it was collected a number of times in the 
sandy river silt along the banks of the Baviaans River. An earlier collection 
by R. Marloth made in 1920 and a recent one by W. J. Hanekom, both from 
the Sutherland Division, indicate that it has a much wider distribution range 
to be investigated. 


434 


Journal of South African Botany 



Fig. 7. 

Lachenalia whitehillensis: 1. Seed side view; 2. Seed from above; 3. Seed from below 

{Barker 1094). 


Specimens Examined 

CAPE PROVINCE — 3220 (Sutherland): Farm Uitkyk Valley, Roggeveld (-AD), 
Oct. 1920, Marloth 9712 (PRE); Hout Hoek, Sutherland (-CA), 13/9/1971, Hanekom 
1574 (PRE). 

—3320 (Montagu): Tweedside (-AB), 25/9/1931, Compton 3960 (BOL), 28/9/1951, 
Barker 9175 (NBG); Whitehill, Laingsburg Division (-BA), 20/9/1943, Compton 
14845 (NBG, holo.), 21/9/1943, Barker 1094 (NBG); Karoo Garden, Whitehill, 
17/9/1945, Compton 17394 (NGB); Karoo near Matjiesfontein, Oct. 1920, Marloth 
9576 (PRE). 

Lachenalia giessii Barker, sp. nov.; habitu nano bulbo tunicis multis bru- 
neis in collum productum, distinguitur. 

Folia duo, variabiles, lineares vel lorata, flaccida, sine maculata. Inflores- 
centia racemosa; pedicelli patentes in longitudem variabiles. Flores pusilli, 
pauci vel multi, albi; segmenta interiora leviter longiora vel segmenta exter- 
na; segmenta tota signis fuscatis prope apices. Stamina leviter exerta. Semen 
oviodeum, arillo minuto terminali, porcato. 

Type Material: South West Africa/Namibia — 2616 (Aus): Garub W of Aus, 
Liideritz District (-CA) 1/9/1963, H. Merxmiiller and W. Giess 3401 (M, iso.; PRE, 
iso.; WIND, holo.). 

Plant 60-160 mm high, usually dwarf. Bulb ovate 10-30 mm diam., cov- 
ered with many layers of brown tunics produced into a neck of fibres 20-50 
mm long, the inner ones darker and cartilaginous, the outer becoming paler, 
softer and often spongy in texture; the membranous leafsheath is often long- 
er than the neck fibres. Leaves 2, without markings, flaccid, and very vari- 
able in shape, from linear, conduplicate to lanceolate-lorate, 2-20 mm 
broad, suberect to spreading, longer than the peduncle. Peduncle usually 
shorter than the inflorescence. Inflorescence racemose, few to many- 
flowered, from 25-40 mm in diam.; pedicels spreading 2-12 mm long, 
elongating in the fruiting stage and spreading at right angles; bracts minute, 


Six new species of Lachenalia (Liliaceae) 


435 



l.i n ) w.i^i 1. ; H MerxmOUerftW.Giena 

• ^ Nr. JQ*1 


Fig. 8. 

Lachenalia giessii: sheet in Botanical Research Institute, Pretoria ( Merxmuller and 

Giess 3041, isotype). 


lanceolate, membranous, white. Flowers 5-8 mm long, narrowly campanu- 
late, tube white, 1,5 mm long, narrowing to the base; outer segments ovate-ob- 
long, subacute, up to 6 mm long and 3 mm broad, white with reddish-purple 


436 


Journal of South African Botany 



Fig. 9. 

Lachenalia giessii: sheet in the Windhoek Herbarium ( Giess Wind. 12845). 

gibbosities; inner segments obovate, a little longer than the outer, white 
with a reddish-purple marking at the apex; stamens slightly exserted, 
spreading; ovary globose; style filiform, finally exserted. Capsule obovate, 
membranous with a few seeds in each loculus; seed globose, aril minutely 
ridged, almost obsolete. 






Fig. 10. 

Lachenalia giessii: 1. Seed side view; 2. Seed from above; 3. Seed from below ( Giess 

Wind. 14656). 


437 


Six new species of Lachenalia (Liliaceae) 


Diagnostic Characters 

Lachenalia giessii is a dwarf species with white campanulate flowers, 
reminiscent of some of the dwarf Ornithogalum species, but each segment 
has a distinct dark marking at the apex. The pedicels which are usually 
shorter than the flowers, often lengthen considerably in the fruiting stage, 
and the leaves which are flaccid and without markings, vary in shape from 
linear, conduplicate to lanceolate. The seed which is globose, belongs to the 
group with an almost obsolete aril. 

The distribution of Lachenalia giessii appears to be confined to the south 
western corner of South West Africa/Namibia between latitudes 26 °-28 °. It 
has not been recorded south of the Orange River, although it has been 
found at Schakalberg (Jakhalsberg) near the northern bank. The earliest re- 
cord was made by S. Karen Regius in 1941, south of Aus, and numerous 
collections have been made in the area since, mainly by Mr. W. Giess, re- 
cently retired Curator of the Windhoek Herbarium, some while on expedi- 
tions with Dr. Merxmuller of Miinchen, West Germany. The species is 
named in honour of Mr. W. Giess in recognition of the valuable work he has 
done in building up the Windhoek Herbarium collections. 

It appears to be most closely allied to L. pearsonii (Glover) Barker, a 
more delicate plant, which is only known from the type collection, which 
was made by Professor H. W. Pearson on the Percy Sladen Memorial Expe- 
dition to the Great Karasberg, South West Africa, 1912-1913. Unfortu- 
nately, no living material has been available for study or comparison of 
either of the two species. The seeds of L. pearsonii are unknown. Its flowers 
are smaller and the segments are described as spreading from the base, and 
it flowers in January. In L. giessii the segments appear to be erect, and it is 
in flower during August and September. 


Specimens Examined 


SOUTH WEST AFRICA/NAMIBIA— 2616 (Aus): Farm Klein Aus, W of Aus, 
Liideritz District (-CA), 9/8/1959, W. Giess and D. v. Vuuren 917 (M, PRE, WIND); 
Garub, W of Aus, Liideritz-Siid, 1/9/1963, H. Merxmuller and W. Giess 3041 (M, 
iso.; PRE, iso.; WIND, holo.); South of Aus (-CB), Sept 1941, S. Karen Regius s.n. 
(Wind); North of Aus, 8/9/1973, W. Giess 12827 (M, PRE, WIND); Farm August- 
felde, 7/9/1973, W. Giess 12805 (M, PRE, WIND); Farm Kubub, 9/9/1973, W. Giess 
12845 (M, PRE, WIND). 

— 2617 (Bethanien): Tirasberge, Bethany Dist. (-BA), 16/8/1963, H. Merxmuller and 


W. Giess 2857 (M, PRE, WIND). „ 

—2716 (Witputs): Farm Witputz-Sud, Liideritz District (DC), 24/9/19/2 h. 
Merxmuller and W. Giess 28776 (M, PRE, WIND); Farm Spitzkop, 15/8/1976, W. 


Giess 14656 (M, PRE, WIND). 

— 2816 (Oranjemund): Schakal Mt. (-BA), 


1/8/1977, M. Muller 781 (WIND). 


438 


Journal of South African Botany 


Lachenalia moniliformis Barker, sp. nov.; bulbo parvo, producentibus 
stolone crassis multis, terminatis bulbilis carnosis, distinguitur. 

Folia teretia viridia; usque ad octo; non amplectentia pedunculi basi. 
Folia singularia propter fascias circulares elevatas per superi duos longitudi- 
nis trientes, monilia simulans; pars basalis fasciata vel maculata, marronines- 
cens prope basin. Inflorescentia racemosa. Flores campanulati, pusilli, opal- 
escens; segmenta interiora vix longiora vel segmenta externa. Stamina “ 
manifeste exerta. Semen ovoideum, arillo minuto terminalis porcato. 

Type Material: Cape Province — 3319 (Worcester): Lemoenpoort (-CD), 

22/9/1978, P. L. Perry 795 (NBG, iso.), cultivated at Worcester Karoo Garden; 
5/9/1979, P. L. Perry 795 (NBG, iso.), cultivated at Compton Herbarium; 6/9/1979, 
P. L. Perry 795 (NBG, holo.), collected in wild habitat. 

Plant 120-170 mm high (up to 330 mm when cultivated). Bulb globose to 
depressed globose, 15 mm diam., covered with a few membranous, light- 
brown scales, producing many bulbils on long stout stolons from the base. 
Leaves many, up to 8, erect to spreading, shorter than the peduncle, 60-160 
mm long (up to 250 mm when cultivated), filiform terete, green with circular 
raised fleshy bands along two-thirds of their length, giving them the appear- 
ance of strings of beads, lower third banded or spotted with dark maroon, 
shading to magenta at the base. Peduncle up to 90 mm long, very slender, 
green, finely and densely spotted with dark maroon. Inflorescence racemose, 
20 mm diam. (25 mm in cultivated specimens), rachis 40-80 mm long (up to 
110 mm in cultivated specimens), finely spotted with maroon; flowers lax, 
cernuous; bracts very small and narrow lanceolate, membranous; pedicels 
3-5 mm long, white, spreading. Flowers 5-7 mm long, 4 mm diam., trun- 
cate at the base, campanulate, tube very short, very pale blue; outer seg- 
ments narrow oblong, 4-6 mm long, extreme apex recurved, very pale blue 
at base, shading to pale pink with reddish-brown gibbosities; inner segments 
only slightly longer than outer, obovate, apex subacute, recurved, white 
with reddish-brown markings near apex; stamens up to 11 mm long, well ex- 
serted; filaments white; ovary green, obovate, 2 mm diam.; style white up to 
11 mm long, finally exserted beyond the stamens. Capsule obovate, mem- 
branous 4 mm diam. ; seeds ovate with very short terminal ridged aril. 

Diagnostic Characters 

The small racemose, campanulate flowers of Lachenalia moniliformis re- 
semble those of several other species in the genus, particularly those of L. 
juncifolia , which also has terete leaves, but these are two in number, while 
in the former species they can number up to eight. Five other species pro- 
duce many leaves, but the beaded structure of the leaves in L. moniliformis 
is quite unique in the genus, and the production of bulbils on long stout sto- 
lons is also of specific importance, while its seeds fall into the group with 
small ridged, terminal arils. 


439 


Six new species of Lachenalia ( Liliaceae ) 



Fig. 11. a, b, c, d. ... 

Lachenalia moniliformis: a. Whole plant; b. Inflorescence; c. Bu Wit u i s 7 qP 
stolons and banded leaf bases; d. Portions of two monihform b a es ( cr 

holotype). 



440 


Journal of South African Botany 



Fig. 12. 

Lachenalia moniliformis: 1. Seed side view; 2. Seed from above; 3. Seed from below 

( Perry 795, holotype). 


The original plant of this completely unknown species, was brought to 
the Compton Herbarium, on 22 September 1978, by Miss P. L. Perry of the 
Worcester Karoo Botanic Garden, and was collected by her at Lemoenpoort 
in the Worcester district. While the flowers were of a familiar type found in 
several species in the genus, the plant had five terete leaves. This many- 
leaved character is only shared by five other species, and in addition these 
are unique in having a structure reminiscent of a string of beads, which sug- 
gested the specific name. 

As only one plant was known, a number of black and white photographs 
were taken, and the inflorescence and several leaves were pressed for the re- 
cord. The plant flowered again on 5 Sept. 1979, this time producing eight 
leaves, which were much longer and more channelled above, near the base, 
and the beading had become less distinct. 

Miss Perry visited the type locality again and was able to obtain a few 
seedlings to cultivate at Worcester. It was then decided to sacrifice the orig- 
inal plant to serve as a type. On lifting the bulb, it was found to be compara- 
tively small, but it had produced a number of stout stolons from its base, 
which were terminated by well-developed bulbils, one of which had already 
developed a single leaf. 

On 7 September 1979, after a very dry season in the area, the original 
site was visited again, and after a search in a somewhat sandy eroded area, a 
few plants were located in flower. Two of these plants were preserved as 
typical examples of wild specimens, and two were planted and cross-polli- 
nated for seed. 


Specimens examined 

CAPE PROVINCE— 3319 (Worcester): Lemoenpoort (-CD), 22/9/1978, P. L. 
Perry 795 (NBG, iso.), cultivated at Worcester Karoo Garden; 5/9/1979, P. L. Perry 
(NBG, iso.), cultivated at Compton Herbarium; 6/9/1979, P. L. Perry 795 (NBG, 
holo.), wild specimens. 


441 


Six new species of Lachenalia (Liliaceae) 

Lachenalia ameliae Barker, sp. nov.; habitu nano distinguitur. 

Bulbus tunicis paucis, pallidis brunneis in collum productum. Folia unum 
vel duo, ovata vel lanceolata; lamina atroviridis, supra glabra vel pilis brevi- 
bus vel longis, interdum limitatus ad margines; pagina inferna marronina; 
basis amplectens, alba, subtiliter marronina punctata. Inflorescentia spicata. 
Flores cremicolores prope basin, palide flavescens tumoribus viridibus in 
segmentis externis, segmenta interiora longiora vel segmenta externa, ebur- 
nea lucentia; apices purpurei tincti. Stamina segmenta intima aequantia. Se- 
men globosus, arillo porcato. 

Type Material: Cape Province— 3320 (Montagu): 5 km S of Bloutorine Station 
(-AD), 7/9/1979, C. Malan 90 (BOL, iso.; NBG, holo.; PRE, iso.). 

Plant dwarf, 40-115 mm high. Bulb globose to depressed-globose, 8-15 
mm diam.; covered with a few membranous light-brown tunics produced 
into a short neck. Leaves 1-2, blade 5-20 mm broad, ovate-acute to lanceo- 
late, spreading to recurved, upper surface smooth or with few to many short 
to long hairs sometimes confined to the margin only, lower surface dark 
maroon, clasping base pale, finely spotted with magenta. Inflorescence 
spicate, 20-80 mm long, peduncle 10-60 mm long, bracts minute; flowers 
12-14 mm long, sessile, spreading, narrow-urceolate, narrowed toward the 
base, tube cream, 4 mm long, outer segments 6 mm long, ovate-acute, 
cream at base shading to pale yellow above, with pale green gibbosities, and 
sometimes with a narrow maroon central stripe, inner segments up td 10 mm 
long, 3-4 mm broad, longer than the outer, the two upper overlapping, all 
cream to pale yellow with a central green keel, apices spreading to recurved, 
occasionally tinged with pale magenta at the apex as in the type collection, 
or with distinct dark purple tips; stamens finally as long as the inner seg- 
ments, decurved; ovary 3 mm long, 2 mm diam., pale green, style cream, 
finally a little longer than the stamens. Capsule membranous, ovoid, 5 mm 
diam.; seeds ovoid with ridged aril of medium length. 

Diagnostic Characters 

Lachenalia ameliae is characterised by its dwarf habit, ovate to lanceo- 
late leaves, dark green on upper surface with or without long or short hairs, 
the lower surface maroon. The inflorescence is spicate, the flowers narrow- 
urceolate, cream and yellow, the inner segments longer, the apices spread- 
ing to recurved, pale magenta to dark purple. 

The earliest discovery of this species was made by Dr. R. Marloth at 
Zwartkoppies near Spes Bona in the Ceres district, in September 1921, and 
the specimens are preserved in the National Herbarium, Pretoria. This was 
followed by a collection by Professor R. H. Compton at Gansfontein in the 


442 


Journal of South African Botany 



Fig. 13. a, b, c. 

Lachenalia ameliae: a. Whole plant {Mauve, Oliver, Malan 206, PRE); b. Left plant 
with smooth leaf, right plant with few marginal hairs; c. Bases of plants with leaves 
with hairs on leaf surfaces ( Malan 90, holotype, NBG). 


Six new species of Lachenalia (Liliaceae) 


443 


Ceres District in August 1935, and another by G. J. Lewis and E. Esterhuy- 
sen about ten miles from Montagu on the road to Ladismith in September 
1935. These are both in the Bolus Herbarium collection. Most of the leaves 
of all these collections are incomplete, but they usually appear to be lanceo- 
late, with a dense covering of short hairs on the upper surface, while the 
flowers have distinct dark purple tips to the inner segments. 

More recent collections were made five kilometres south of Bloutoring 
Station, first by Mrs. A. A. Mauve, accompanied by I. Oliver and C. Malan, 
in September 1974, when three living specimens were made available for 
study with some ripe seeds. The same locality was visited in September 1979 
by C. Malan and W. F. Barker, when it was possible to study the living 
plants in their natural habitat. They were found growing in red sandy clay 
on windswept flats, usually on humps with small karoo plants holding the 
soil. The shape of the leaves in this area proved to vary considerably, from 
broadly ovate to lanceolate, and the upper surface of the dark green blade 
was either quite smooth or the integument consisted of hairs of varying 
length and density covering the surface, or confined to the margins only. 

None of the flowers examined in this locality had dark purple tips to the in- 
ner segments, but in some cases the extreme tips were tinged with magenta. 
The seeds available in both groups were similar in character, with small 
ridged arils. 


Fig. 14. 

Lachenalia ameliae: 1. Seed side view; 2. Seed from above; 3. Seed from be ow 
{Malan 90, holotype, NBG). 

The species has been named in honour of Mrs. A. A. Mauve who 
brought the first living specimens from the type locality to my notice, and 
has contributed so much to our knowledge of the Liliaceae. 

Specimens Examined 



2 


3 


9/1974, A. A. Mauve, I. Oliver & C. Malan zoo (.indvjj, j ™ ^ D 



444 


Journal of South African Botany 


tion, 7/9/1979, C. Malan 90 (BOL, iso.; NBG, holo.; PRE, iso.); About ten miles 
from Montagu along old road to Ladismith, Sept. 1935 (fruiting), Lewis & Esterhuy- 
sen (BOL). 

Acknowledgements 

I am much indebted to Dr. J. P. Rourke for the Latin translations of the 
diagnoses; to Miss P. L. Perry for bringing L. moniliformis to my notice, 
and for taking me to the type locality, and also to Miss C. Malan for arrang- 
ing an expedition to the type locality of L. ameliae. 


J1 S. Afr. Bot. 49 (4): 445-449 (1983) 


STUDIES IN ORCHIDACEAE FROM SOUTHERN AFRICA 

Graham Williamson 

(P.O. Box 499, Oranjemund, South West Africa/ Namibia) 

Abstract 

Habenaria lithophila Schltr. subsp. mossii Williamson subsp. nov. is described. 
Nervilia humilis Schltr., a new record for Southern Africa, is illustrated. 


UlTTREKSEL 

STUDIES VAN ORCHIDACEAE VANAF SUIDER-AFRIKA 

Habenaria lithophila Schltr. subsp. mossii Williamson subsp. nov. word beskryf. 
Nervilia humilis Schltr., ’n nuwe aantekening vir Suider-Afrika, word geillustreer. 

Key words: Habenaria, sp. nov., Nervilia, new record, Orchidaceae, South West 
Africa/Namibia. 

Habenaria (sect. Diphyllae ) lithophila Schltr. subsp. mossii Williamson, 
subsp. nov. a subspecie typica petalis lobis anticis et labello lobis lateralibus 
teretibus, calcari multo longiore distinguenda (Fig. 1). 

Typus: South Africa, C. E. Moss 14412 (J, holo.). 

An erect terrestrial herb 180-240 mm high. Stem slender to stout, 
straight, 2-5 mm in diameter, rising from an ellipsoid tuber up to 20 mm 
long and 15 mm in diameter. Basal leaves 2, opposite, prostrate, ±orbicular, 
acute to cuspidate, longest up to 70 mm long and 60 mm broad; cauline 
leaves up to 23, erect, clasping stem, linear-lanceolate, acute, up to 15 mm 
long. Inflorescence densely up to 40-flowered, 70-90 mm long and 25 mm in 
diameter; flower bracts up to 17 mm long; flowers greenish, suberect; ovary 
with pedicel lightly curved, up to 15 mm long. Intermediate sepal erect, nar- 
rowly ovate, acute, ca. 7 mm long and 3 mm broad; laterals deflexed over 
ovary, longitudinally obliquely ovate to elliptical, subacute, ca. 8 mm long 
and 4 mm broad. Petals bipartite almost to the base; posterior lobe erect, 
adnate to the dorsal sepal, linear-ligulate, margin ciliolate, ca. 6 mm long 
and 1,2 mm broad; anterior lobe curved upwards horn-like, terete, densely 
ciliolate, acute, ca. 8 mm long and 0,6 mm in diameter. Labellu n tripartite 


Accepted for publication 2nd June, 1983. 


445 


446 


Journal of South African Botany 


with side lobes curved upwards and middle lobe curved forwards and down- 
wards; middle lobe ligulate with ciliolate margins, ca. 8 mm long and 1,2 
mm broad; side lobes terete, densely ciliolate, subacute, ca. 9 mm long and 
0,8 mm in diameter; spur cylindrical, pendent to horizontal with a narrow 
base and slightly inflated apex, up to 23 mm long. Anthers erect, rounded, 
ca. 2 mm high, canals porrect to slightly curved upwards, ca. 1,8 mm long; 
stigmatic arms porrect with flat upper surface and rounded lower surface, 
apices obtuse, ca. 2 mm long; rostellum centre lobe triangular with obtuse 
apex curved backwards, ca. 1,2 mm high and much shorter than anther. 

Habitat: open grassland on dolomite or in black sandy soil. 

The new subspecies differs from the typical subspecies in that the ante- 
rior petal lobe and labellum side lobes are more or less terete and the spur is 
very much longer, measuring up to 24 mm. 

This new species is named in honour of Prof. C. E. Moss who, in 1917, 
was appointed the first Professor of Botany at the University of the Wit- 
watersrand. He collected extensively on the Witwatersrand (Gunn and 
Codd, 1981). 

TRANSVAAL: Witwatersrand, Johannesburg, Thorntree Kloof, April 1927, C. E. 
Moss 14412 (J, holotype); 2 miles WNW of Krugersdorp, in grassland on dolomite, 
March 1954, A. O. D. Mogg J 29146 (J); W Rand, Sterkfontein, amongst “Olifants- 
klip” on the knoll, March 1954, H. B. Gilliland s.n. (J); Johannesburg, Waterval in 
black sandy soil, March 1957, J. E. Bartlett J 30224 (J). 

Nervilia humilis Schltr. in Bot. Jb. 53: 551 (1915). 

This new record was made by Mr. S. Venter in the company of Mr. D. 
Hardy in March 1977 while on a collecting trip to Swaziland. The plant colo- 
nies consisted of up to 30 plants growing in deep humus in shaded submon- 
tane woodland. The plant illustrated (Fig. 2) flowered in cultivation in the 
nursery of the Botanical Research Institute in February 1981. 

SWAZILAND: Piggs Peak at ± 1 300 m, col. March 1977, fl. Feb. 1981, Hardy and 
Venter 4186 (BOL, PRE). 

Acknowledgements 

The author is grateful to the curators of the Bolus Herbarium, University 
of Cape Town, the Kew Herbarium, Royal Botanic Gardens, Kew and the 
Moss Herbarium, University of the Witwatersrand, Johannesburg, for use of 
their facilities and the loan of material. 

Thanks are also due to Dr. P. J. Cribb, Kew Herbarium, Royal Botanic 
Gardens, Kew, for advice and guidance, Messrs D. Hardy and S. Venter for 
material of Nervilia humilis and to Prof. E. A. Schelpe, Bolus Herbarium, 


Studies in Orchidaceae from southern Africa 


447 



Fig. 1. 

Habenaria lithophila subsp. mossii. A, habit, x f; B, flower, x 4; C, dorsal sepal, 
x 2; D, lateral sepal, x 2; E, petal, x 2; F, labellum, x 2; G, column, x 8, H, 
rostellum, x 6. All from A. O. D. Mogg J 29146 (J). All magnifications are approxi- 
mate. 


448 


Journal of South African Botany 



Fig. 2. 

Nervilia humilis. A, habit, X 1; B, leaf, X 1; C, dorsal sepal, 
x 1; D, lateral sepal, x 1; E, petal, X 1; F, lip flattened, X 6; G, part of column, 
x 10; H, anther cap, x 10. From Hardy and Venter 4108 ex Hort B.R.I. All magni- 
fications are approximate. 


Studies in Orchidaceae from southern Africa 449 

University of Cape Town, for guidance, advice and for reading the manu- 
script. 

References 

Gunn, M. and Codd, L. E., 1981. Botanical Exploration of Southern Africa: 
254-255. Cape Town: Balkema. 

Schlechter, R., 1915. Orchidaceae Stolzianae, ein Beitrag zur Orchideenkunde des 
Nyassa-Landes. Bot. Jb. 53-504. 



J1 S. Afr. Bot. 49 (4): 451-454 (1983) 


A NEW SPECIES OF ERIOSEMA (FABACEAE) FROM THE EASTERN 
TRANSVAAL 

Charles Stirton* 

(Botanical Research Institute, Private BagXIOl, Pretoria 0001, R.S.A.) 
Abstract 

Eriosema naviculare C. H. Stirton, is described as new. This rare species is con- 
fined to the boundaries of the Lowveld and Lowveld Sour Bushveld vegetation types 
in the Pretoriuskop-Numbi region of the eastern Transvaal. 

UlTTREKSEL 

’N NUWE ERIOSEMA (FABACEAE) SOORT VANAF OOS-TRANSVAAL 
Eriosema naviculare C. H. Stirton word beskryf. Hierdie skaars soort is beperk 
tot die Laeveld en Laeveld-Suurbosveld plantegroei tipes in die Pretoriuskop-Numbi 
streek van Oos-Transvaal. 


Key words: Eriosema, sp. nov., Fabaceae, eastern Transvaal. 

Eriosema naviculare C. H. Stirton, sp. nov.; E. saligno E. Mey. proxi- 
mum, sed foliis latioribus et floribus minoribus differt. 

Planta perennis usque 500 mm alta. Folia trifoliolata. Foliola naviculifor- 
mia subtus coacta glandulosaque. Inflorescentia 30-40-floris. Flores 10-13 
mm longi, bractea 6 mm longa, 1,5 mm lata, anguste lanceolata caduca. Ca- 
lyx 6 mm longa, dentibus aequalibus breviter triangularibus pubescentibus, 
glandulis parvis albis interspersis. Vexillum orbiculare, flavum, signo nec- 
tario viridi, appendiculis bifidis auriculos carentibus. Alae flavae carinam 
pallidiflavam superantes. Fructus 20 mm longi, 11 mm lati, molliter alba pu- 
bescentes. 

Type: TRANSVAAL— 2531 (Komatipoort): 5 km from Hazyview to Numbi 
Gate (-AA), Stirton 9829 (PRE, holotype). 

Erect perennial herb up to 500 mm tall arising from a clavate rootstock 
with 1-several lateral branches. Stems 1-5, minutely clothed in white hairs. 
Stipules 16 mm long, 4, 5-5,0 mm wide, free, lanceolate-falcate, senescent 
before leaf expansion. Leaves trifoliolate, except for the 30-40 mm long, 


* Present address: The Herbarium, Royal Botanic Gardens, Eew, Richmond, 
Surrey TW9 3AE, England. 

Accepted for publication 11th July, 1983. 


451 


452 


Journal of South African Botany 


20-28 mm wide, unifoliolate leaves produced at the first two nodes; terminal 
leaflet 72-100 mm long, 41-63 mm wide, elliptic, symmetrical, stipellate; 
laterals smaller, asymmetrical with width ratio 1:2,3; all leaflets felt-like with 
raised veins below, covered in minute raised glands; strongly boat-shaped; 
petiolules 4-5 mm long, 3 mm wide; rhachis deeply channelled. Inflores- 
cence 36-40-flowered, axillary, 210-240 mm long at maturity, 40 mm long 
when first flower opens. Flowers 10-13 mm long, each subtended by a 6 mm 
long, 1,5 mm wide, narrowly lanceolate, rapidly caducous bract. Calyx 6 
mm long; tube 3 mm long; teeth equal, shortly triangular, 3 mm long, finely 
pubescent with small white glands interspersed. Standard 11 mm long, 8-9 
mm wide, emarginate, orbicular, yellow with green nectar-guide, back finely 
pubescent with many small white glands; appendages fused, bifid, hooded, 
free from auricles; claw 5 mm long. Wing petals 10 mm long, 5 mm wide, 
yellow with a few hairs along the base; claw curved, 2 mm long. Keel blades 
8 mm long, 3 mm wide, pale yellow, indented at base, with few hairs scat- 
tered along the lower margin, outer face with pale glands towards the apex; 
claw upcurving, 2 mm long. Androecium 8 mm long; vexillar stamen free, 7 
mm long, geniculate. Pistil 7 mm long; ovary 3 mm long, white pubescent; 
style glabrous, height of curvature 2,25 mm, thickened at point of flexure; 
stigma minutely capitate. Fruits 20 mm long, 11 mm wide when green, softly 
white pubescent; beak straight, 1,5 mm long. Seeds unknown. 

Considering the lack of access roads in the general area of distribution it 
is not surprising that this distinctive species has not been collected more of- 
ten. E. naviculare can be confused with E. salignum E.Mey. (yellow- 
flowered form) but is readily separated from that species by its broad pubes- 
cent leaflets and much smaller flowers. 

The specific epithet, “naviculare”, meaning boat-shaped, refers to the 
distinctive manner in which the leaves are borne from their axis. 


Specimens Examined 

TRANSVAAL — 2531 (Komatipoort): 5 km from Hazyview to Numbi (-AA), 
Stirton 9829 (PRE); Shabani, van der Schijff 1394 (PRE, K); Pretoriuskop (-AB), 
van der Schijff 1498 (PRE, PRU, K), Acocks 16622 (K). 


Distribution and Biology 

Eriosema naviculare is restricted to grassveld in the Pretoriuskop-Numbi 
area. It has only been collected three times before. In all cases it was found 
in burnt veld. Flowering occurs from late October to early January. 

As with all Eriosema in South Africa, nothing is known about the biology 
of this species. This genus should be investigated as all the South African 
species are densely glandular on most organs (against phytophagous insects) 


New species of Eriosema (Fabaceae) 


453 



Holotype 


S.ZJ. 


&c& L e HER BAR 

JM 

PRETORIA 

2531 AA 


| *«,- TRANSVAAL 

CH Stirton 

9839 n* 

* 

| A 1981. 1(1. 23 

[£ 

Eriosema aviculare C.H. 

Sti 

rton Sf. nov. 

5 km from Hazy view to '< 
Songani Hospital ) 

Burnt, red sandy loam, 
Soots tock vertical with 

umt>i Gate (Just past 
Sushveld 

laterals; Flowers 

— 




Fig. l 

Holotype of Eriosema naviculare C. H. Stirton (voucher is Stirion 9829, not 9839 as 
marked on the sheet). 


454 


Journal of South African Botany 


and secondly have fleshy funicular remnants on the seed; the latter are cer- 
tainly connected with ant dispersal of the explosively dispersed seed. 

Diagnostic Features 

Erect herb; clavate rootstock; naviculate trifoliolate leaves; free, senes- 
cent lanceolate-falcate stipules; yellow flowers; caducous flower bract; equal 
and shortly triangular calyx teeth; standard with bifid hooded appendages 
free from the auricles and wing petals exceeding the keel blades. 

Acknowledgements 

I would like to thank Miss C. M. Wilmot-Dear for the Latin diagnosis; 
and Mrs. Jana Zantovska Stirton and Mr. B. Schrire for accompanying me 
in the field during my search for this species. Thanks also go to Dr. B. de 
Winter, Director, Botanical Research Institute, Pretoria, for arranging for a 
photograph to be taken of the type and for giving permission for it to be re- 
produced here. 


J1 S. Afr. Bot. 49 (4): 455-460 (1983) 


Book Reviews 

Dictionary of Cultivated Plants and Their Regions of Diversity: Ex- 
cluding Most Ornamentals, Forest Trees and Lower Plants, by A. C. 
Zeven and J. M. J. de Wet, with pp. 263 and 110 maps. 2nd rev. ed. Wage- 
ningen: Centre for Agricultural Publishing and Documentation (Pudoc), 
1982. Hard cover Dfl. 90,00 (R47 plus tax). 


Reviewers of books featured in this Journal are asked to be constructive. In the 
present instance this is very difficult for as far as I can see its only commendable fea- 
ture is overall appearance; size, pagination and format are in accordance with the 
best traditions of Pudoc. The binding seems tough and durable and is a great im- 
provement on that of the original edition. There are a number of printing errors — 
Sanseverinia for Sansevieria and Euphrobiaceae for Euphorbiaceae are examples — 
and these are probably the result of slap-dash proofreading. But as far as the con- 
tents are concerned, for the life of me I cannot understand how this book achieved a 
second printing! I can only assume either that purchasers are extraordinarily ignorant 
of current concepts in taxonomy and phytogeography or that there is a chronic 
dearth of reference works in this field. And to add insult to injury, it has doubled in 
price. 

The first edition was reviewed by the writer in 1976 [// S. Afr. Bot. 42 (2)] in 
which he described the general layout of contents. The work has now been enlarged 
to include about 2 300 species; unfortunately it has also incorporated a greater num- 
ber of errors despite the comment of the senior author that mistakes and omissions 
have been rectified. The co-operation of a new co-author (Professor P. M. Zhukovs- 
ky died in 1975) does not seem to have benefited the present edition. 

In the Preface it is stated that taxonomy is based primarily on Willis’s dictionary 
(1966) and the Bailey Hortorium, Hortus Third (1976). Chromosome numbers are 
derived from Bolkhovskisch et al. (1969). Airy-Shaw revised Willis in 1973 and a far 
better conspectus of chromosome numbers was published under the auspices of the 
Missouri Botanical Garden in 1981. This explains the appallingly out-dated nomen- 
clature and a plethora of inaccurate author citations; I counted 42 in the African Re- 


gion section alone. 

The drawings are hardly designed to aid taxonomy, for those on pages 99, 1UU, 
111, 134, 136, 173 would not pass muster in a Botany I practical class; that of Sorg- 
hum bicolor looks like the sweeping-broom a certain Dutch admiral used to hoist to 
his mast-head. Malus prunifolia has its origin in Northern China and East Asia and 
yet is included in the Near Eastern Region. Prunus bessyi belongs to North America 
but its drawing is firmly set down in the Chinese-Japanese Region. This is very mis 
leading. Many species are asterisked but it is left to the imagination to reso ve e 
meaning of this symbolism. In fact it indicates that there is a related reference else- 
where in the book. „ , „ ? .• ,. r 

The aim of this work is to give the reader quick reference to the regions ot diver 
sity of cultivated plants and related wild species. If so, why is peach listed as ha g 
secondary centre of distribution in California and not in Georgia. - couth 
not been included in the U.S. A.? A number of plants now widely cultivated inj 
ern Africa are not included whereas some of dubious value receive 


455 


456 


Journal of South African Botany 


palathus linearis and Cyclopia genestoides are not mentioned and the taxonomy of 
the buchus is hopelessly inaccurate. 

Distribution maps do not inspire confidence and the snippets of data given in the 
text are inadequate. Dioscorea hirtiflora for example, occurs in dense forests as well 
as “Savanna” and Oxytenanthera abyssinica extends to the limits of the tropical zone. 
The maps depicting Central American and Mexican Region and North American Re- 
gion exclude the West Coast of the U.S.A. completely. Presumably the authors are 
waiting for a cataclysmic shift of the San Andreas Fault to raft California out to sea, 
whereupon it can be treated as an island ecosystem! 

I apologise to Pudoc but not to the authors for this scathing review. They ought 
really to have done their homework more thoroughly. 


O. Kerfoot 


The Study of Plant Structure: Principles and Selected Methods, by T. 
P. O’Brien and M. E. McCully, with pp. x -I- 248. Melbourne: Termarcar- 
phi Pty Ltd, 1981. Distributed outside Australia by Blackwell Scientific Pub- 
lications, approx. US $32,00. ISBN 0 9594 174 0 0. 

Books on plant microtechnique do not appear frequently. All anatomists and 
morphologists — in fact all botanists seriously interested in plant structure — are famil- 
iar with Johansen’s classic Plant Microtechnique (1940) and Sass’ Botanical Micro- 
technique (3rd edition, 1958). These two books were followed by Jensen's Botanical 
Histochemistry (1962), a pioneer in its field but now crying out for revision and up- 
dating (an immense task!), as well as Botanical Microtechnique and Cytochemistry by 
Berlyn and Miksche (1976), purportedly the successor to the Sass book, but some- 
what heavy and disappointing. 

One of the outstanding botanical papers of 1968 was that by N. Feder and T. P. 
O’Brien which appeared in the American Journal of Botany (55: 123-142) under the 
title “Plant microtechnique: some principles and new methods”. This paper intro- 
duced the revolution which subsequently occurred in plant microtechnique by pro- 
viding convincing evidence that there were available methods far superior to the clas- 
sic coagulative fixative-paraffin wax embedment techniques. 

Now O’Brien (Monash University, Australia) and Margaret McCully (Carleton 
University, Canada) have, in the book under review, produced a worthy successor to 
Johansen’s, Sass’ and Jensen’s books. The Study of Plant Structure will be welcomed 
by the microtomist for, in addition to many novel methods, there are a host of practi- 
cal tips and an excellent list of references. 

The book is organised into six chapters. The first is a short presentation of physi- 
cal principles as they apply mainly to the perception of structure, while the second 
considers methods of microscopy and includes valuable information on fluorescence, 
a hitherto much neglected aspect of botanical microscopical procedure. Chapters 3 
and 4, respectively, deal with morphological and anatomical methods. Chapter 5 is 
on photomicrography. Finally, Chapter 6, the practical heart of the book, encom- 
passes what the authors refer to as Formulary. As the term implies, this is a glossary 
of protocols involving everything from replicas, whole mounts and macerates to fixa- 
tives, embedment, sectioning and staining. The book is concluded with 11 appen- 
dices and a long list of references, but — most unfortunately — with no index. This 
omission, together with a weak binding, is the book’s most serious defect. 


Book Reviews 


457 


The authors might have argued that the more than five pages of contents negates 
the preparation of an index, but finding one’s way around can be quite frustrating. It 
is not made any easier by the bureaucratic blur of digits. Example: Negative staining 
is listed as 6.15.13.6 under 6.15.13— Staining Methods for Electron Microscopy, and 
found on p. 6.102. There are also other — but minor — irritations such as a large num- 
ber of typographical errors (example: James and Jape instead of James and Jope) 
and the mentioning of obviously interesting and specific methods (example: fluores- 
cent staining of DNA by 4'6-diamidino-2-phenylindole) without giving the full pro- 
cedures. 

The authors have produced a mine of well-documented and pertinent informa- 
tion, ranging from careful, stepwise procedures such as, for example, the re-purifica- 
tion of glutaraldehyde to valuable hints on the physical manipulation of materials, 
techniques and instruments. Their explanations of the principles of microscopy, fixa- 
tion and staining are uncomplicated and lucid. The history and development of plant 
microtechnique is also adequately covered by their extensive coverage of the relevant 
literature. 

If ever so slight an impression might be gained that O’Brien and McCully are ac- 
tually writing for O'Brien and McCully, it should be seen against the background of 
the vast collective microtechnical experience of both of these pioneering - and, it 
should be added, excellent — investigators. The Study of Plant Structure is a highly 
recommendable compendium. 


Molecular Biology of Plant Tumours, edited by G. Kahl and J. S. Schell, 
with pp. xxiv + 615. New York, London, Paris, San Diego, San Francisco, 
Sao Paulo, Sydney, Tokyo, London: Academic Press, 1982. US$69,50. 
ISBN 0 12 394380 9. 


Chris H. Bornman 



search: Van Montagu, Schell, Chilton. 
Part 3, other potential vectors such as 
zobium are also considered. 


458 


Journal of South African Botany 


'*^This book is timely, coming as it does at a period when “molecular plant genet- 
ical engineering” has became a catch phrase, but yet underscoring the point that man 
is able to learn his lessons from Nature. As Schell likes to point out, Nature, in the 
form of Agrobacterium tumefaciens , has set a perfect example of how genes are en- 
gineered, and moved, from micro- to macro-organism. Applying these lessons, a 
number of laboratories (such as Ghent, Versailles, Leiden, Cetus-Berkeley) have 
shown that the integration of foreign DNA into the plant cell genome is not only 
possible, but that its transmittance through meiosis to subsequent plant generations 
is feasible. 

Although comprised of many chapters by more than twice that many authors, the 
book is well edited and supplied with an extensive index. The papers on the whole 
are eminently readable and well-illustrated, and should serve as valuable sources of 
reference. It is not only the molecular biologist who will benefit by this book, but tis- 
sue culturalists generally, regardless of whether their primary interest is genetics, 
biochemistry or plant physiology. 


Chris H. Bornman 

Physiological Plant Ecology: II. Water Relations and Carbon Assimi- 
lation, edited by O. L. Lange, P. S. Nobel, C. B. Osmond and H. Ziegler, 
with pp. xi 4- 747 and 153 figures. New York, Heidelberg, Berlin: Springer- 
Verlag, 1982. Volume 12 Part B in the New Series “Encyclopedia of Plant 
Ecology”. DM 288, approx. US$120,10. ISBN 3 540 10906 4. 

This is the second in the four-part series on physiological plant ecology. It is a 
large book (747 pages) made up of 18 chapters plus an Introduction, an author in- 
dex, a taxonomic index and a subject index (which is comprehensive and very use- 
ful). 

The title gives equal emphasis to water relations and carbon assimilation, and 
although the two are inextricably linked in an ecological sense (as emphasised by the 
authors in their Introduction), there is in fact far more weight given to water re- 
lations. Fourteen of the chapters are concerned primarily or exclusively with water, 
and only four with carbon assimilation. 

The book begins with a useful overview of water in the soil-plant-atmosphere 
continuum. The next four chapters deal with water status, uptake, movement and 
storage in plants, followed by two chapters on water loss (via plant surfaces and via 
stomata). All of these chapters are detailed, up-to-date and of a high standard. 
There is then a chapter which reviews the various mathematical models of water loss. 
The chapter itself is descriptive with virtually no mathematics of its own. 

The following three chapters provide a thorough, physiological and physical ac- 
count of plant water stress. They provide a most useful and interesting account of 
this subject. 

Following two chapters on germination (of seeds and spores) and one on flood- 
ing, the final four chapters deal with photosynthesis and carbon-water relations. The 
chapter by Osmond, Winter and Ziegler on the functional significance of different 
pathways of CO 2 fixation in photosynthesis is of particular note. The first half of the 
chapter is a well-presented review of the fundamental biochemical basis of the three 
photosynthetic CO 2 fixation pathways. It is followed by sections on the physiological 
context and the ecological context, and the chapter is recommended reading for all 
who are interested in the physiological basis of plant ecology. The final chapter, by 
E.-D. Schulze, on plant life forms and their carbon, water and nutrient relations, is 


Book Reviews 


459 


habits o7plants efUl ° f ^ physiolo § ical basis for different forms and growth 

Technically the book is well produced. There are virtually no typographical er- 
rors. Figures and tables are clear and the cross-referencing and common style suggest 
u dltmg ' Consid i enn .g today’s book prices, the cost is not outrageous (a mere 
$120) but it is a pity that it will be beyond the means of most individuals, since the 
book deserves to be m the personal library of all plant ecologists. It is a must for all 
libraries concerned with botany, ecology and agriculture. 


B. H. Walker 


Advances in Research and Technology of Seeds: Part 6, edited by J. R. 
Thomson, with pp. 124. Wageningen: Centre for Agricultural Publishing 
and Documentation (Pudoc), 1981. Dfl. 36,00. ISBN 90-220-0786-3. 

This is again a worthwhile contribution in which recent advances in four perti- 
nent topics in the field of the physiology of seed development and germination are 
reviewed. The following topics are covered: Ecological aspects of seed germination 
(P. A. Thompson); Physiological and biochemical events in seed development 
(Daphne J. Osborne); Changes in seed composition during germination (J. C. 
Slaughter); and Germination of seeds (A. Lovato). 

Although all four reviews are valuable in assisting the seed scientist to keep in- 
formed of new developments, the review by Thompson is an approach from a differ- 
ent angle and particularly interesting subjects such as maintenance of plant popu- 
lation, germination strategies and environment and phenotypic plasticity of 
germination responses in relation to geographical distribution are discussed. 

Advances in Research and Technology of Seeds is primarily intended for seed re- 
search workers and seed technologists. Apart from the seed specialist, however, this 
volume would also be of special interest to plant ecologists in South Africa. 

Plant names and subject matter are well indexed. 

J. G. C. Small 


Advances in Research and Technology of Seeds: Part 7, edited by J. R. 
Thomson, with pp. 140. Wageningen: Centre for Agricultural Publishing 
and Documentation (Pudoc), 1982. Dfl. 36,00. ISBN 90-220-0802-9. 

This volume contains four reviews primarily covering papers published after 
1977. 

Weed seed investigations by R. J. Chancellor reviews dormancy, periodicity of 
germination, seeds in the soil, survival of buried seed and seedling emergence. The 
possibilities and need for further investigations on the use of chemicals for breaking 
weed seed dormancy as an effective control measure are clearly shown. 

Effects of radiation on seeds is aptly reviewed by A. D. McKelvie. In addition to 
ionising radiation effects, in which field most papers were published, the review men- 
tions a few papers on microwaves, ultrasound and magnetic fields. 

The shift in research emphasis from mutation induction to stimulation of growth 
and yield by radiation is interesting. 

The review by D. L. Smith on seed development and structure emphasises mor- 
phology. This is welcomed in a science dominated by physiology and biochemistry 
Seed identification remains a problem for seed analysts and it is of interest to note 
the possible use of the computer as a tool in seed identification. 


460 


Journal of South African Botany 


Research on environmental and genetic factors affecting tree and shrub seeds is 
reviewed by B. S. P. Wang, J. S. Pitel and D. P. Webb. 

Although papers on dormancy dominate this review, other interesting topics such 
as effect of pesticides and seed mass on germination, seed dispersion and seed losses 
due to insects, rodents and disease, are covered. 

This volume, with a good subject index, should appeal to a wide range of plant 
research scientists including geneticists, foresters and ecologists. 

J. G. C. Small 


ANNOUNCEMENT 

It is a great pleasure to announce the founding of the International Association 
for Landscape Ecology. The objectives of this organisation are to enhance communi- 
cation among scientists, planners and managers in landscape ecology, and to pro- 
mote the development of knowledge in this field and its dissemination and applica- 
tion. 

The landscape concept deals with areas on a scale which may contain many rela- 
tively homogeneous ecosystems, or landscape elements, which invariably differ in the 
degree of human activity or impact associated with them. Landscape ecology, there- 
fore, studies the structure, function and development of landscapes. Structure in this 
context may refer to the spatial dimension and distribution of landscape elements; 
function to the flow of energy, mineral nutrients and species among these elements, 
and development to the change in the landscape as a whole through time. 

An interest in the development of concepts in this field, or the application of 
such concepts, forms the kernel of the International Association, whose members in- 
clude, inter alia , ecologists, geographers, planners, landscape architects, foresters, 
wildlife biologists, land managers, urban and regional designers and environmental 
scientists. Landscape ecology is, of course, a formal discipline in a number of Euro- 
pean countries. 

Anyone who has an interest in landscape ecology is invited to send their name 
and address to: — 

O. Kerfoot 

Regional Representative for IALE, 

Department of Botany, 

University of the Witwatersrand, 

1 Jan Smuts Avenue, 

JOHANNESBURG 2001, 

R.S.A. 

You will receive the first issue of the IALE Bulletin and other announcements 
once you are on the mailing list. 


JOURNAL OF SOUTH AFRICAN BOTANY 


VOLUME 49 


1983 


INDEX OF PLANT NAMES 


Note. Plant names appearing in tables or lists are not included in this index. Page references to 
new taxa are printed in bold type. An asterisk indicates an illustration. 


Page 

Boophane flava Barker ex Snijman 

243 , *245, *246, *247 

Bowiea gariepensis E. J. van Jaarsveld 
343 , *345 

Eriosema naviculare C. H. Stirton 

451 , *453 

Euphorbia namuskluftensis Leach 

189 , *191, *192 

Freylinia decurrens E. J. van Jaarsveld 


61, *62 

F. visseri E. J. van Jaarsveld 57, *60 

Galium rourkei Puff 185, *186 

Habenaria lithophila Schltr. subsp. 
mossii Williamson 445, *447 

Kyllinga alata Nees 280, *283 

K. alba Nees 284, *288 

K. albiceps (Ridley) Rendle 274 

K. brevifolia Rottb 287, *290 

K. elatior Kunth 290, *293 

K. erecta Schumach 292, *296 

K. intricata Chermezon 292, *294 

K. melanosperma Nees 296, *299 

K. nemoralis (J. R. and G. Forster) 

Dandy 272, *275 

K. odorata Vahl 278, *281 

K. pauciflora Ridley ’ 268, *270 

K. polyphylla Willd. ex Kunth . . . 298, *301 

K. pulchella Kunth 271, *273 

K. welwitschii Ridley 276, *278 

Lachenalia aloides (L.f.) Hort. ex 
Asch. & Graev. var 48 

L. ameliae Barker 441, *442 

L. bachmannii Bak 49 

L. bolusii Barker 51 


L. bulbifera (Cyrillo) Hort. ex Asch. 

& Graeb 

L. camosaBak 

L. concordiana Schltr. ex Barker 

L. contaminata Ait 

L. elegans Barker var 51, 


Page 


Lachenalia fistulosa Bak 47 

L. framesii Barker 50, 54 

L. giessii Barker 434, *435, *436 

L. hirtaThunb 48, 54 

L. juncifolia Bak 46, 47 

L. longibracteata Phillips 53 

L. maximiliani Schltr. ex Barker 51 

L. moniliformis Barker 437, *439 

L. montana Schltr. ex Barker 51 

L. mutabilis Sweet var 49, 50 

L. namaquensis Schltr. ex Barker 55 

L. nordenstamii Barker 428, *430 

L. pallida Ait 48 

L. polypodantha Schltr. ex Barker .... 55 

L. pusilla Jacq 52 

L. reflexaThunb 47 

L. rosea Andr 49, 52 

L. rubidaJacq 52 

L. salteri Barker 52 

L. schlechteri Bak 48, 54 

L. splendida Diels 49 

L. trichophylla Bak 48, 50 

L. undulata Masson 49 

L. unicolor Jacq. var 48, 50 

L. unicolor Jacq. var. fragrans (Jacq.) 

Bak 51 

L. unifolia Jacq 47 

L. unifolia Jacq. var. wrightii Bak 47 

L. ventricosa Schltr. ex Barker 50, 53 

L. violacea Jacq 53 

L. whitehillensis Barker 432, *433 

L. zebrina Barker 424, *424, *425 

L. zebrina Barker forma densiflora 

Barker 426 

Leucospermum hamatum Rourke 


Linum acuticarpum Rogers 183 

L. africanumL '83 

L. comptonii Rogers 183 

L. gracile Planch 183 

L. heterostylum Rogers 183 


Mimetes hirtus (L.) Salisb. ex Knight 

125, *128 


461 


462 


Index to Plant Names 


Page 

Nervilia humilis Schltr 446, *448 

Otholobium pictum C.H. Stirton 340, *341 

O. rubicundum C.H. Stirton 337, *339 

Polyxenasp 52 

Psoralea implexa C.H. Stirton . . . 329, *330 

P. trullata C.H. Stirton 332, *334 

Tritonia atrorubens (N.E.Br.) L. Bol. 400 

T. bakeri Klatt 354 

T. bakeri Klatt subsp. bakeri .... *335, 357 

T. bakeri Klatt subsp. lilacina (L.Bol.) 

De Vos 358 

T. chrysantha Fourcade 387, *388 

T. delpierrei De Vos 403, *405 

T. disticha (Klatt) Bak 375 

T. disticha (Klatt) Bak. subsp. disticha 

377 

T. disticha (Klatt) Bak. subsp. rubro- 

lucens (Forster) De Vos 378 

T. drake nsbergensis De Vos 397 

T. flabellifolia (Delaroche) G.J. Fewis 

363 

T. flabellifolia (Delaroche) G.J. Lewis 

var. flabellifolia 366 

T. flabellifolia (Delaroche) G.J. Lewis 
var. major (Ker-Gawl.) De Vos 

, *364, 367 

T. flabellifolia (Delaroche) G.J. Lewis 


var. thomasiae De Vos 368 

T. flava (Ait.) Ker-Gawl 412 

T. florentiae (Marl.) Goldbl 410, *411 

T. fulvaDehnh 413 

T. karooica De Vos 407, *408 

T. laxifolia (Klatt) Benth. ex Bak. 


T. lineata (Salisb.) Ker-Gawl 369 

T. lineata (Salisb.) Ker-Gawl. var lin- 
eata 370, *372 

T. lineata (Salisb.) Ker-Gawl. var. 

parvifolia De Vos 374 

T. magnifloraDehnh 413 

T. marlothii De Vos *405, 406 

T. moggii Oberm 401, *402 

T. nelsonii Bak 398 


Page 


Tritonia pallida Ker-Gawl 359, *360 

T. pallida Ker-Gawl. subsp. pallida . . . 361 

T. pallida Ker-Gawl. subsp. taylorae 

(L.Bol.) DeVos 362 

T. parvula N.E.Br 389 

T. pulchella Dehnh 413 

T. securigera (Ait.) Ker-Gawl .... 384, *385 
T. strictifolia (Klatt) Benth. ex Klatt . . 390 

T. tripunctata Dehnh 413 

T. tugwelliae L. Bol 419, *420 

T. undulata (N.L. Burm.) Bak 413 

T. watermeyeri L.Bol 395, *396 

Tulbaghia acutiloba Harv *254 

T. alliacea L.f *253 

T. capensisL *253 

T. cernua Ave- Lall *253 

T. coddii Vosa & Burbridge *255 

T. cominsii Vosa *256 

T. dregeana Kunth *253 

T. galpinii Schl : . . . . *254 

T. leucantha Bak *255 

T. ludwigiana Harv *253 

T. macrocarpa Vosa *254 

T. montana Vosa *255 

T. natalensis Bak *255 

T. nutans Vosa *254 

T. simmleri Beauv *254 

T. tenuior Krause & pinter *253 

T. transvaalensis Vosa *254 

T. verdoornia Vosa & Burbridge *255 

T. violacea Harv *256 

T. violacea Harv. var. maritima Vosa *256 
Tylecodon kritzingeri E.J. van Jaars- 

veld 305, *308 

Xanthoparmelia aggregata Knox 144, *146 

X. cedrus-montana Brusse 145, *147 

X. cylindriloba Knox 147, *149 

X. dysprosa Brusse et Knox 148, *149 

X. exornata (Zahlbr.) Brusse et Knox 

150, *151 

X. fucina Knox 150, *153 

X. hedbergii Knox 152, *153 

X. ianthina Brusse 152, *156 

X. karoo Knox et Brusse 154, *156 

X. leucostigma Brusse 155, *157 

X. walteri Knox 157, *158